The best parts of the Official Unofficial 2012 VMware Storage Survey are the open ended questions – some of the things people put in there are really interesting, funny, eye-opening, heck sometimes offensive :-)
The second open-ended question in the 2012 survey was “What are the things you WISH we would do?” – and indeed – I got hundreds of examples. Pretty well every vendor got a couple slaps in this list. These are going RIGHT to the product teams (warning, some firm language)… Read on past the break and judge for yourself…
What are cool things you wish EMC (or other storage vendors) would do?
- *cough* include multipathing software for free *cough*
- - Stretched clusters for small ESX clusters (2-4 ESX hosts) - Reduce SSD price
- 40gbit fc
- A online lab environment. So people like myself can get in depth hands on experience
- A plugin to manage san / nas from vSphere client, like Dell vCenter management plugin.
- A tool that did online block reallocation for VM's
- A virtual appliance like SANsymphony 7.0 Storage Virtualization
- APIs
- Ability to manually configure RAID level in EMC VNXe
- Add FC to Isilon!!
- Adopt ZFS
- Advertise prices instead of making it hard to find
- Affordable storage
- After the VNXe, which can make any IT guy into a storage expert, anything is possible.
- Allow data to be rebalanced accross a disk pool when you add drives to expand it.
- Allow me to move data (on a lun basis) from one frame to another without selling my soul.
- An optimizer calculator to achieve 80% fill of vmfs
- Anything to make CIFS backups faster and more easily recoverable.
- Application aware snapshots of virtual machines for replication.
- Application level consistancy for vm snapshots for things like SQL, Exchange, etc.
- Arrays that integrate with dedupe devices as a tier for very old data.
- As i storage guy, maybe more integration tool with vmware vcenter
- Atomic-level storage - 20YB sitting in a small box on my desk.
- Auto profiling and give recommendations on raid sets.
- Auto tiering, play better with VM backup's
- Auto-tiering, VAAI assists, Thin Provisioning
- Automate tiering
- Automatic aligning. Best practice guide(s). Easier guides for "standard" setup.
- Automating capacity planning based on constant inputs and mathematical formula.
- BC/DR in a multi-tenant model for vCD environments would be awesome. :-)
- Be able to automatically find VMs with performance problems based on disk issues.
- Be able to back up and restore VMs within a vCD enviroment.
- Be more open with there source drivers to all Virtualization Operating Systems (VMware, Xen, etc)
- Be quicker to support SRAs for mirrorview... what the hell EMC
- Beter report manager tools.
- Better GUI, with less reliance on CLI
- Better Integration of RecoverPoint with VNX for File (or Celerra) and VAAI support with vplex
- Better VASA implementations
- Better application recovery for VMs (referring to HP/3PAR here)
- Better automation, better integration with VMware environments (if it is possible)
- Better backups of RDM's integrated with snapshots
- Better bundles for enterprise customers that include packages like Operations, SRM, etc
- Better communication, documentation, training
- Better explain their VMware integration
- Better integration and more things to help with multi-data center environments.
- Better integration and sanity checks - e.g. warn on double-thin provisioning.
- Better integration between block & file.
- Better integration of NAS and better monitoring tools that integrate with vCenter.
- Better integration of toolsets. Challenging in a fast moving market I know.
- Better integration with all the VMware cool stuff
- Better integration, easier to use and monitor. Free training.
- Better integration, management and storage performance in vSphere
- Better interoperability in terms of SAN replication.
- Better perfoming object based storage
- Better performance tools
- Better replication options for WAN's
- Better reporting. Navisphere is like dumb
- Better storage management tools INCLUDED with the hardware
- Better support
- Better test facilities for us to build PoC's in their cloud.
- Better vCenter Integration
- Better vSphere integration
- Better vSphere integration for queisced snapshots and replication.
- Better visibility into why VMs are running slow and more initiators in the storage frames.
- Better vmware/storage self provisioning
- Better, widely supported virtualization(storage)
- Better/more plugins to help bridge the gap between SAN administrators and VMware architects.
- Bi-Directional array based replication for ESX stretch cluster across sync + async distances.
- Big Partys :D
- Block based de-dup (where is it guys?)
- Block level Dedupe, our NX4 only does it for NSF shares.
- Block-level deduplication on the vnx
- Books I can find on Amazon etc.
- Bring a supported EMC virtual storage appliance to market.
- Bring on the virtual appliances; "real", grid-capable Avamar Virtual Edition, VPLEX vApp.
- Bring storage closer to the compute layer
- Build servers to sell all the solution
- Bundle products free with VMware licenses? :)
- Bundle storage management with the hardware (one price gets all), while staying competitive.
- Buy SUSE from Attachemate.
- Buy Veeam and integrate with there own products
- COnsolidate storage and backup products.
- Can't really think of anything right now....
- Can't think of any for now.
- Can't think of any right now.
- Can't think of anything other than give me a free VNXe ;)
- Catch up with Netapp software offerings
- Charge much less!
- Cheap & Integrated Cloud Storage
- Cheap affordable storage
- Cheaper SSD
- Cheaper Solutions
- Cheaper manteinance contracts
- Cheaper storage options :)
- Clearer visibility of what space is being used where and how.
- Closer VM/Storage management
- Closer integration with ESX, better and easier performance measurment in SAN
- Come up with standard terminology across vendors.
- Compete with EqualLogic on price so that customers of ours could start small and grow with EMC
- Complete integration with VMware products
- Compression. Got 93% on a win2008 image.
- Context aware computing and big data (pattern search)
- Continue brining enterprise features to the SMB market with SMB IT budgets in mind
- Continue to bring down the cost of storage.
- Continue to improve integration VMware - vendor. Crate vendor independent solutions.
- Continue to offer plugins to integrate storage management into the vSphere client.
- Continue to provide information and transparency to the community.
- Continue to provide tighter integration via plugins.
- Cost less!
- Create more Virtual Appliance for Testing Arrays.
- Create trending tools
- Cross platform between storage vendors
- Cut costs.
- Cylon lights on storage array.
- Data Domain, OINIX
- Decent non-Indian support.
- Dedup
- Dedup, integration with other vendors
- Dedupe on live data
- Dedupe primary highperforce san
- Deduplication at any level
- Deduplication everywhere and fully automated
- Deduplication in VNX block
- Deduplication over several boxes
- Deduplication similar to netapp.
- Deduplication! Integration with backup software (e.g., NetBackup) for off-host backups!
- Dell EQL - wish they'd have a synchronous replication option.
- Destroy networker, which is the bane of my life ;-)
- Detecting bottlenecks
- Die, Die, Die
- Direct attached SAS Tape Libraries
- Disaster recovery at the VM level, instead the LUN level.
- Discovering new storage in better and more stable way
- Dispense with the thick protocols and go over PCIe.
- Distibuted seamless storage
- Distributed power management for storage arrays.
- Ditch Navisphere
- Do Metro Cluster configuration easy
- Do more for the SMB market.
- Doing lots of cool things, just want to be able to take advantage
- Don't know
- Don't know at the moment
- Donate to vTrusrcommunity
- Dont know EMC products
- Drop prices :-)
- Drop the price of VPLEX. :)
- Drop their prices for hardware.
- EMC - Tailor to SMB market more. (Price, features, licensing) Official VSA product.
- EMC VNX should become truly unified -- no need to provision disks for file only.
- EMC is already cool
- EMC is cool enough already. But more coolness will be welcomed!
- EMC is not cool, too pricy.
- EMC is thebest so far for me.
- EMC storage is very good and flexible
- Ease new lun deployment
- Ease the provisioning of storage, lun's security, backup and virus checking.
- Easier to implement and manage
- Easy Replication Plugin for Store complete vmware Infrastructure in a Thing linke "Snapshot"
- Easy VSAs
- Easy storage expansion / scaling
- Easy thin space reclamation
- Easy to sell tiering and deduplication for virtualized environments.
- Easy upgrade paths
- Easy ways to isolate issues from the VM-Host-HBA-FC Switch-Array-LUN
- Eliminate the need for VMware SRM with replication and easy HA/Long Distance VMotion.
- Encryption at rest!
- End to End VM awareness.
- Enhance VASA. Allow tiering at the application level (VM). Fail over at the VM level.
- Enhanced array based backups Backup products for vCloud
- Even more integration with hypervizors
- Everyone loves SWAG :)
- Fairly happy with the direction EMC is going right now.
- Faster zoning/LUN presentation, with the same SCSI ids presented to all hosts in a hostgroup.
- File level deduplication.
- Flash Cache for Write IOPS.
- Focus more on decentralised storage architectures.
- For us would be trying to market something a bit lower cost for EDU/Gov usage
- Free SRM type products
- Free hardware for all techies :-p
- Free training and certification for partners
- Full Flash Storage
- Full cooperation of Storage DRS with integrated storage balancing mechanism at the EMC devices.
- Full integration vasa/vaai
- Full vsphere client integration
- Full-blown powershell CLI to the storage array
- Fully integrate to the VMWare api's like UCS is doing. Plugins to the management.
- Further integration with the hypervisor. Making the storagesolution easier.
- Further integration with virtual platforms
- Further reduce windows dependencies via appliances
- General introduce vendor independent standards for storage.
- Generally make storage easier to understand... especially EMC storage!
- Get a real management & orchestration product for Vblock.
- Get better usage of FAST and Pools. Some features are only marketing stuff.
- Get cheeper for the SME's
- Get hot vMotion supported.
- Get stretched clustering RIGHT.
- Give away free demo arrays (SATA is fine).
- Give away free things.
- Give beer away with storage,
- Give me EFD and FAST on VNXe
- Give me free stuff :)
- Give me storage and training
- Give me storage arrays for free.
- Give me tool to find hotspot on harddrives/Luns as a customer.
- Global Dedupe
- HA mechanisms across data centers >40ms latency to help support more/better DR mechanisms.
- HP to support SSD in MSA / P2000 G3
- Handling storage on VM level
- Hands on Demo's of equipment
- Have a common standard for array level replication between vendors.
- Have an easy to understand building/spec tool for hosts and SAN storage.
- Have fixed pricing.
- Have more SMB solutions. They act live every business should have $50000 to spend.
- Have more free training.
- Have more immediate knowledge of interfacing with VMware environments.
- Have more lunch-n-learns/presentations/seminars in Victoria BC. :-)
- Have my vspeciaist spend more time with my account!!!
- Help our storage team
- Hire me!
- Hire people like me
- Honestly, EMC's product line is too complicated.
- Hummm can't think of any right now
- Hurry it up with Project Lightning! :)
- Hurry up with the granular availability shown in VSP1700 at VMworld! Super cool stuff!
- Hybrid drvies
- I am happy with whats on offer at the moment.
- I am not THAT advanced with storage.
- I do not use EMC for VM.
- I don't know anything about EMC. We have just started using VMware
- I don't know very well all the EMC products
- I like the way that the VASA is headed.
- I think EMC is already doing amazing things, what magic will they think up next !
- I think EMC is on the right track for the SMB with the VNXe.
- I think storage vendors are doing a good job of advancing the technologies.
- I want VMware to show guests in storage/SAN. Like NPIV.
- I want beat MetroCluster! Maybe with somthing like VPLEX/SE.
- I want the vVolumes that were previewed at VMworld NOW! :)
- I want to see FAST on the VNXe
- I wish 3PAR's vCenter plug-in was improved.
- I wish HP would bring back SVSP.
- I wish all vendors would adopt the Compellent licensing model.
- I wish other storage vendors would go away
- I wish the gui was a little more intuitive and responsive
- I wish there were better tools/wizards etc to setup, test and verify storage multipath stuff.
- I wish they would work more closely with us. I'm not feeling the 'love' from EMC as a partner.
- I would like to see more automated storage tiering in conjuction with vmware storage drs
- I would like to see when a LUN is thin inside of vcenter
- I would love it if EMC would spend some time making EMC Connect more user friendly.
- I'm pretty much happy with how it goes now.
- IPv6 needs a good implementation IMHO. We have lots of customers wanting to try it.
- If i could think of something awesome i'd probibly be working for them
- Implement more granular FAST Auto Tiering and realtime for the VNX.
- Improve software stability (windows clients) - Netapp Snapdrive bugs have been a major issue
- Improved management and monitoring UI.
- In all seriousness - I think the VNX really nailed it as far as feature sets and wish lists.
- Increase mean-time between disk failures, Applicaiton aware RecoverPoint replicaiton.
- Instead of licensing for different features emc should have a cost for all features
- Integrate automated teired storage with SSD's onto the lower end offerings.
- Integrate better with Backup Vendors (Netbackup)
- Integrate further with VMware
- Integrate performance and size metrics into vCenter.
- Integrate srm with vplex!
- Integrate with other vendors for replication, etc.
- Integrating VAAI and VASA out of the box, currently is just a PITA
- Integration of products. Example: Avamar, VNX, Recover Point, etc.
- Integration with VMware at virtual machine level, make the storage a vm-aware system.
- Interact better with VMware
- Intergrated management for all platforms in the environment.
- Intregration, K.I.S.S. :)
- Introduce new products asap
- It'd be nice if EMC storage was easier to manage. Unisphere is still a steaming pile.
- Just continue all uber coolness, i like it
- Just keep doing it...EMC is cool
- Just keep me update with all the cool stuff. EMC always stays ahead of the game.
- Keep driving prices down.
- Keep going and inventing together with vmware
- Keep on developing the VPLEX-thingy - but please improve the report-generating part of it
- Keep on doing what you are doing
- Keep up the good work in making stable products with EASY and RESPONSIVE support.
- Keep up with NetApp Block level dedupe,cache aware dedupe, ...
- Lay out the options easier, low cost hardware
- Less complicated and have one pain of glass. Fit in to the overall monitoring scheme.
- Local roadshow with hands on labs.
- Love the VNXe...wish EMC would give me more control over raid groups.
- Love the idea of metro clustering
- Low Tier Deduplication (7k SATA/SAS Tier) NAS using ZFS
- Low price VEEAM type product
- Lower Costs of basic storage.
- Lower Prices
- Lower barrier of entry for the cool stuff!
- Lower cost :)
- Lower cost IP replication storage arrays
- Lower cost! Licensing is a nightmare
- Lower costs of hardware and standardize network connectivity options.
- Lower pices :)
- Lower price$ :) :)
- Lower prices
- Lower prices for higher ed customers.
- Lower prices.
- Lower prices... I'm new to the whole VM world so i don't have an opinion yet.
- Lower the cost while maintaing performance.
- Lower the cost/TB
- Lower the costs of storage; making storage array upgrades easier then they are now.
- Lower the power consumption.
- Lower the price point :D
- Lower the prices? Give us access to more technical documentation of Isilon
- Lower their prices LOL
- Lowering prices :)
- Make DR easier - keep replication terminology consistent.
- Make I/O stats easy to access per file system or LUN.
- Make Project Lightning happen soon.
- Make SAN replication easier and more reliable.
- Make Thin LUNs space reclaimation easier/automated.
- Make VMFS aware of real storage consumption on the array
- Make a single pane of glass interface for non vBlock customers.
- Make active-active affordable?
- Make advanced storage features cheaper!!
- Make affordable iSCSI storage options with dual controllers for SMB market.
- Make all the features compatible
- Make array firmware upgrades easier.
- Make cheaper shared storage solutions.
- Make deduplication and thin provisioning a free feature on as many storage products as possible.
- Make documentation more like VMware.. EASY!
- Make easier to generate reports that deal with thin provisioning better
- Make iscsi easier to implement
- Make it easier to locate performance issues.
- Make it easier to trouble shoot LUN trespassing.
- Make it easier to understand performance issues at the vm guest level.
- Make migrations between arrays easier
- Make monitoring array performance easier
- Make monitoring storage easier / prettier
- Make more entry level devices
- Make more perfoamnce data available with vcenter
- Make my life as the vm admin easier
- Make out life even easier
- Make reporting easier
- Make setup and configuration easier.
- Make storage cost less... and for more features not get nickeled and dimed to death.
- Make storage performance = compute performance.
- Make storage performance problems easier to troubleshoot.
- Make storage predictions easier, based on trends of past storage growth.
- Make storage provisioning easier for VM admins.
- Make stuff easier... Storage SHOULD be easy...
- Make the mapping easier with the datastores and the underlining luns
- Make things a little more affordable!!!
- Make upgrades easier
- Metro clusters
- Metro clusters without vplex
- Metro-clusters are cool.
- Metroclustering our Celerra NS-960. Integrating Rainfinity into Celerra/Unisphere.
- Modular server a la Intel
- More SMB targeted products with enterprise features VNXe is great for this!
- More Tech roadshows displaying the tech in use not just marketing fluff.
- More advanced thin provisioning
- More affordable performance
- More array vAppliance emulators for POC and Training purposes
- More basic training courses for who may not be able to afford it
- More consistency across VNX/VNXe range in terms of protocols offered and underlying software.
- More education on integrating the products and getting the most performance and efficiency.
- More entry-level shared storage devices
- More feasible solutions for smaller enterprises.
- More flash based possibilities. Etc
- More free items, at least for the community. Keep them coming!!
- More granular view of vdisks.
- More integration into vmware. More flexibility.
- More integration of storage in vsphere.
- More integration to reduce the number of management interfaces required.
- More integration with VMware and Cloud
- More integration with VMware technologies...
- More integration with View and thin provisioning.
- More integration with vCenter around reporting (performance and capacity).
- More integration with vCenter. Storage performance information readily available there EASILY.
- More options for SMB clients.
- More plugins and intergration with vSphere
- More plugins to connect the storage managment with vCenter
- More power to the Vmware admin :-)
- More price accessible Storage Virtualization Inter compatibility
- More reference build with server vendors
- More storage arrray control from within vSphere Client
- More storage solutions for SMB. An I mean EMEA SMB, not USA SMB :-)
- More stuff for homelabs where you get good performance...
- More tiered storage
- More training
- More userfriendly API / CLI interfaces
- More vSphere integration ONE INTERFACE to manage ALL..lol
- More vaai integation
- More videos on the web to talk about the EMC products and features.
- More virtual appliances and support for said appliances for non production use.
- Move FAST to the VNXe, integrate WAN accelleration into the SAN.
- Move object-based storage down to lower tiers of products (I hate RAID).
- Multi-vendor array based replication.
- Multiple replication partners (eql), easier license features, just all in one box.
- N/A
- N/a
- NA
- NFS
- NFS (EMC) Other vendors - better replication options to remote site.
- NFSv4 Sessions
- NVRam Storage?
- Native or wizard driven install inside VMware components. Dedup with flat license model
- New form factor
- No complaints other than support from our existing storage vendor (HP)
- None
- None really
- None.
- Not an expert as to imagine cool things EMC doesn't make as today, sorry.
- Not at the moment.
- Not much
- Not much else
- Not sure on this (yet)...
- Nothing
- Nothing Special
- Nothing comes to mind.
- Nothing in particular
- Nothing really at the moment. I'm not on the VMAX team, so I don't know what they're up to.
- Nothing special
- Offer Power Shell integration.
- Offer an inexpensive expandable unit to support as little as 3 hosts.
- Offer cheaper version for home labs!
- Offer lab equipment for home use at reduced costs ;)
- Offer more partnerships to VMware partners.
- Official Support of vplex Geo on VMware
- On SAN dedupe!
- On single view for each VM: VM -> Host -> Datastore -> LUN -> Array and vice versa!
- Open standard for server side flash caching (think Project Lightning with HBA vendor buy-in
- Open up the new performance troubleshooting tool to customers!
- Out-of-the-box best practices vCenter Orchestrator tie-ins
- PCIe flash
- Parallel nfs with vmkernel integration
- Partially fill disk shelves
- Pay per use storage models.
- Performance management, time keeping
- Please VPLEX NAS support....
- Please continue to integrate with the hypervisor!
- Please, include a real auto-tiering mixed on Compellent and Netapp tech.
- Plugins for vcenter and means of easily identifying perfomance issues
- Point in Time SRM SRA Adapter for Recoverpoint
- Pools with different Raidlevels
- Powershell intergration
- Pretty happy with NetApp.
- Pretty happy with it at the moment
- Price reduction.
- Primary storage de-duplication
- Pro-sumer devices for home labs
- Production VSA application
- Production ready VSA
- Provide a better status/monitoring interface
- Provide a cheap cost effective storage solution thats cheap and affordable.
- Provide a holistic view from VM through to the SAN to the underlying storage in one tool.
- Provide a smaller business solution for very important data, but caps at 1TB...
- Provide affordable small business (<75 nodes) products
- Provide arrays to Partners for lab use.
- Provide better vSphere Client plugins for storage appliances.
- Provide cheaper SSD so i can replace the spindles
- Provide end user tools to do array sizing
- Provide even better integration with vcenter -one pane of glass please
- Provide granular per-VM, per-Datastore real-time storage performance displays and trending
- Provide more intuitive sizing tools.
- Provide more training with the arrays to configure and operate.
- Provide much more visibility into how the storage is being used, be vm aware (like Tintri).
- Provide the same sort of vCenter integration as NetApp.
- Provide you with performance analysis on a yearly basis for their equipment (free of charge). :)
- Provider better systems/block of storage to make increasing storage easy!
- Put Self Encrypting Drives in the SAN
- Put more tier 1 cache on VNX/Clariion
- Quality SOHO storage
- Re-thin for guest OS
- Recover point to have better VAAI compatibility, in particular when using VNX splitter
- Reduce cost :)
- Reduce cost of flash/SSD solutions
- Reduce cost of hardware
- Reduce costs and improvr features!!
- Reduce the prices!
- Release Analyzer Helper. Would love to play with it!
- Release more products for small businesses.
- Release orchestrator plugin
- Release virtual arrays for testing that anyone can use without first buying the array. Ie, netapp
- Reliability on the storage has been an issue this year (2011 has been a bad year here for EMC!)
- Remote Replication
- Replica and cloud management in vsphere clietn
- Replication of local Disks for SMB
- Replication on a VM basis!!
- S/DRS integration with VASA.
- SAN prices should be less expensive
- SAN with auto-tiering and thick-LUN capability.
- SDRS for SRM. cheaper ssd.
- SRM with MirrorView
- Scale Out Block Storage
- Scale out storage pNFS VAAI and VASA support
- Sell only SSD so then we can continue to ignore disk I/O but no longer worry about it? :)
- Send me swag !
- Sheaper san (nas) iscsi
- Simple backup solutions that earn confidence. DRaaS possibly.
- Simpler and more cost accessible VSA like appliance
- Simplify administration (especially CLI).
- Simplify and standardize the storage API
- Simplify licensing.
- Simplify management, although to be fair HP pretty good
- Simplify mangement Provide better support quicker (allow advanced users to skip tier 1)
- Single management platform.
- Size locked arrays for learning
- Small Business VPLEX
- Smart alerting. To be able to customize alerting so that we do not receive 3000 alerts.
- Snapshot/replication storage
- So many cool things at EMC lately, I'm out of ideas!
- Something about the alignment bewetten VM disks and the storage would be nice!
- Stable GUI FAST for VP (between 10/15k FC)
- Standarization like vBlocks you're already doing
- Standarized
- Start bundling some cool technologies together - inexpensively.
- Stay innovative!
- Still learning current stuff, no time to think about other cool things :)
- Stop going above the technical folks head to upper management...
- Stop licensing every single piece of software.
- Storage DRS offloading on the unified storage. NFS offloading.
- Storage Replication
- Storage Virtualization improvements à la xsigo
- Storage as a Service
- Storage management standardisation
- Storage virtualization. HDS way.
- Stretch NFS
- Support labs with cheaper pricing.
- Swap the Celerra with Isilon in the VNX. :)
- Synchronous / Asynchronous file replication to live filesytem target
- TIghter vCenter integration
- Take over LAN management so I could sack all these ciscophiles and their complications...
- Take us out for dinner more often.
- Talk strategy, improve management tools.
- Tap more into the virtual applicance sector for VMware.
- Tell management how to better desgin storage and vmware teams. :)
- That HP Eva works better with Vmware such ad NetApp
- They already do but we don't have a big budget for it.
- They are already doing cool things
- Think of the little guy more
- Tighter direct connect into Cisco UCS
- Tighter integration (Transparency) with other technology stacks
- Tighter integration with entire infrastructure solution
- Tintri like products. ie VM oriented SAN.
- To be honest I'm happy with the features that our CX4-240 provides.
- To decrease the prices of the appliances ;)
- To have better performance and utilization reports included with the box.
- To make FCOE really work
- Trying to get file server data sent to disk and to the cloud... no more tape!
- Update older hardware to suppor VAAI
- Use smaller block sizes in FASt or EasyTier LUNs, like DELL Compellent.
- Using HP EVA Hoping for vaai support
- Using Netapp withy NFS
- Using SSDs in a hybrid caching form.
- VA based storage may be cool (like VSA, or an unlocked full featured NetApp simulator)
- VAAI
- VAAI NFS for vCloud Director
- VAAI is Cool, Stretch Cluster is Cool. Backup integration in vCD is way Cooler ;)
- VAAI, SVC sctrech cluster
- VASA is it, or at least the start. Its already going in a good direction.
- VDI optimization by using flash for read/write (like Whiptail)
- VIBE
- VM aware storage.
- VM level backups
- VM-aware Storage (a.k.a.VM Volumes)!!!
- VMWare integration, one GUI. Performance. Price.
- VMware on Isilon and you can kill Nutanix in one shot
- VNX with cheap sata drives
- VPLEX SRA
- Vendor-agnostic storage replication without absurdly expensive intermediary hardware.
- Virtual Storage box for testing in an OTA environment
- Virtual appliances for SAN with larger feature set.
- Virtualization on storage arrays
- Vm level diagnostics instead of lun level.
- Vnxe 3300 <-> vnx 5300 remote replication over iscsi
- We are moving away from emc completely.
- We are planning to start using EMC in our environment.
- We can't even do what is currently out there, so wishing for new things would be pretty funny :)
- We need to catchup on VAAI etc first, but we're getting there.
- We use NetApp and are very happy with their solution set.
- We will be checking VPLEX.
- Why all storage doesn't do intelligent storage tiering is beyond me....
- Wider range of products (lab based)
- Wish Isilon could do fiber
- Wish NetApp was more proactive in the blogging scene like the EMC vSpecialists
- Work on increasing storage densities and decreased pricing for enterprise(SLC) SSD
- Work on their partner websites to make it easier to find information
- Work with VMware to build location aware storage in VASA.
- Work with VMware to over come issues with a cluster across multiple datacenters
- Wouldn't know.
- You're doing the things I want. Unfortunately, I'm stuck with IBM unless I switch projects.
- a synchronous storage-replica within 500 km
- access to EMC support content without an EMC support contract
- adanced free training
- an active/active storage cluster.
- auto align VMs
- auto-provisining luns to vmware when more storage is needed.
- automate the host configuration of array in VMware with simple tool
- automatic availability over differents infrastructures
- be cheaper ;)
- better Storage monitoring / configuration within VMware vSphere
- better VMware integration better performance on VNXe
- better VMware integration for all vendors
- better VMware interation
- better compression
- better customer support (EMC)
- better faster CHEAPER
- better integration with VMware products end-to-end
- better powerlink
- better pricing for higher education ;)
- better reporting
- better reporting of storage thru vcenter plugins
- better support better price
- better support staff
- better training but lower in prize
- better vcenter integration - so i dont have to rely on the storage admin for info regarding SAN.
- better vmware integration
- block deduplication
- block level dedupe on VNXe pervasive EFD more memory cache
- bring down 15k scsi costs
- bring out a fibre channel VNXE
- can't think of anything.
- cheap enclosures - ala drobo
- cheap snapshots and replications
- cheap storage
- cheap, high performance, data deduplication on production storage.
- cheaper basic storages ? .)
- cheaper ssd
- cheaper storage
- compliancy check on the storage from within vsphere client
- continued unification of vnx
- create LUN's per VM from in vCenter when you create a VM
- cut back on the licencing of very little thing
- data teleportation
- dedup on block automatic restriping when adding disks to a pool
- dedup, replication
- dedupe in primary storage
- dedupe primary data
- deduplication of iSCSI / FC storage
- deduplication, long distance sync replication
- disclosing true cost of ownership
- drop a zero off the price of their storage
- easy management
- easy up use of dedup and storage virt. for entry levels (smbs)
- end to end mapping
- faster, more, and bigger. easier storage replication integration in vmware srm.
- federate software
- free training on basic storage principles, it is still a black box to the majority of people
- full ssd array's
- geo-replication
- georeplication like gluster & streaping inside the client
- get SSD SANs when it is reliable enough :D
- get certain revs of solaris clustering working with vplex ;)
- get my wife crazy about storage!!!!!!!!!!!!!! YEAH!!!!!!!!!!!!! :)
- get there reporting tool to be integrated with the VMware one's
- give NFR device and licenses to partners
- give a free low cost iSCSI SAN for free.
- give me a better inside what i need for VMware View best performance
- give me a vnxe 3300 ;)
- give me free stuff to study/play with
- give out "features" without extra high pricing
- give out powerpath free.
- give us more beer :))
- going forward to NFS V 4 or similar technologies
- good vnx configuration utility
- gui that is easy to manage / configure
- i can`t think of any in this moment
- i wish there's a better ways to manage LUNS without having too many LUNS.
- if I knew, I wouldn't tell you!
- improve geo-repl..sync,async cost and performance
- improve management, provisioning, replication, recovery, performance.
- inline dedupe on primary storage, writable snapshots
- integrate SRM and View
- integrate better with isilon storage
- integrated management application with all kind of storage
- iphone app
- keep as cool as they are :-)
- less firmware updates, so that i can keep my vmware up to date :-)
- leverage existing infrastructure similar to what the VSA allows.
- looking at recover point & avamar
- low cost 100% SSD Storage
- lower cost of storage, especially licences!
- lower costs
- lower prices + more affordable SSD
- lower the price! Make it more afforadable
- lower the prices of SSD disks
- lower their prices lol
- lower their prices.
- lower their prices. :-)
- make coffee
- make cross-array vendor replication simpler and more common
- make it cheaper, make sure upgrades are easier
- make storage easier and all vendors to follow one standard
- make their units less expensive.
- make things cheaper :)
- make vmware aware of Netapp thinprovision ..i know it is on roadmap
- making long distance clustering easier
- making remote replication cheaper
- mayor integration with VMware API
- mini onboard FC switch
- modular scalability of compute and storage
- more UBER software for testing before we use it in production
- more VAAI features on the lower end arrays
- more afordable
- more and more vaai features
- more balanced comparisons. apples and apples.
- more comprehensive administrative/management tools
- more demos
- more lun management from vcenter
- more make advance features such as Fast easy to use. and include in the price :)
- more options for SMB market
- more perfomance indidators/counters regarding storage/queueing in vcenter. Cheaper SSD
- more tiering
- more visibility from storage side into vmware, and vice versa
- n/a
- na
- no idea
- no one.
- none
- none uniform access solution (VPLEX) supporting asynchronous distances w/ VMware
- none.
- none. cant wait for ssd price to continue free falling
- not much official EMC simulator available out there.
- not sure
- not sure right now. been a very long day
- not sure, dont have EMC
- not sure.
- nothing
- nothing - if we select VNX with recoverpoint all bases are covered
- offer realistic disk prices
- on line vmax labs for customers
- online backup
- pNFS Object Storage Project Lightning
- pay attention to the midsized business, We buy stuff too.
- performance monitoring
- performancemonitoring - where is the bottleneck etc. direct from the storge
- policy based mgmt and end to end QoS -integrated for VM from VM to array....
- powershell cmndlets for their array management
- present lun over fc and iscsi
- produce more reliable software
- provide highest IOPS for VDI projects
- provide inexpensive or included DR software (you have to pay for the remote storage)
- provide more smb solutions at a more realistic price
- provide their own vmotion and svmotion plugins
- put down the prices, so the cool features will be available for the smaller companys
- recplication
- release SRA's sooner
- release a free and full functional virtual appliance for testing the storage capabilities
- revise the pricing model.
- run a hypervisor within the storage, eg vm's inside the vmax
- send me some SWAG!
- simplify, simplify, simplify
- spill to public cloud storage
- ssd caching in lower end models.
- stable, cheap, performant storage systems
- standardize nfs
- stop trying to lock us in, more open modular storage systems, IE use your own disks etc.
- storage de-dupe for VMs
- storage deduplication
- storage provisioning from vCenter Server
- stretched cluster available for cheap arrays
- sub 10k iscsi san w/ snapshots/replication and dual storage controllers
- support VPLEX Geo in VMware Implementation
- the main cool thing, what they can do - low the prices!
- trans dedup with fc, metro mirror 100km sync
- unify management tools
- unknown
- using default Hardware Parts, nothing self made
- vMotion, snapshot should be done at storage level to avoid the VC operations.
- vPlex GEO.. Can't wait... EMCCA keep educating this is really good stuff!!
- vaai for vnxe
- variable block size deduplication on both block and file
- vcenter plugins like netapp has
- virtual storage
- vplex sounds cool
- work out a SSD roadmap that makes sense
- zero-detect thin provisioning (like 3par)
- I like being able to provision my own storage. I like vmware's proposed revamp of storage getting down to file/VM provisioning level instead of LUN level. There was a VMworld session on this.
- Integrate active/passive LUN fail-back within the vSphere GUI. For example to restore your LUN ownership after a Flare code upgrade.
- Ensure that engineers also make it to EMC sales events. Internal engineers like me need someone to talk to.
- Simplify the packaging: I never know exactly just what features I'm actually buying and neither do the sales agents, which says a lot.
- -)vcenter plugins for ISILON -)Can you please create a better or more user friendly documentation website for your products? Have used Powerlink for a few years and it's still horrible....
- Not undercut our sales by offering the same storage for cheaper to our clients just to get the sale direct even though we are partners.
- Come up with a product offering that allows storage to easily scale out instead of up. This product should use commodity hardware and stripe data redundantly across storage nodes.
- continue to simplify array management. VNXe is very simple and my customers like it. VNX is on the right track, but just isn't there yet. Also, upgrades of the VNX can be smoother if you ask me. I'm a EMC partner implementation specialist, and run into several irritating things when upgrading (like no dual CS upgrades using USM , so we have to get EMC support to do it). Keep those best practices coming!
- Automation is #1. Integration with powercli or maybe orchestrator/director and of course vcenter. Everybody we ask about automation whether it be from emc or hp seems to have an incomplete solution. Vsi from emc is nice but limited. Monitoring is #2. An end to end view (graphical would be nice as we're lowly windows admins) with historical base lines and trending. And site to site comparisons would be cool as well, I.e. is my Virginia storage performing better than my San Jose storage and why.
- Don't have EMC at the moment. Current storage vendors are slow to release firmware fully supporting the latest VMware storage features.
- Probably more specific to IBM storage but IP Metro/global mirroring as it all seems to need dark fibre links. Better performance reporting, capacity management.
- Plugins to show performance graphs related to the VMs, tie tighter/automate with vCenter actions.
- Provide access to full documentation and management consoles on their websites, without requiring a subscription/support contract.
- End to end performance of VM to VNX including MDS FC Directors. Hope ProSpere gets there soon. ECC is awfull and Navi Analyzer is not to far behind. Trying to analyze 14 NAR files to look at a weeks worth of performance is painfull and a month would cause me to pry my eyes out with a SAS connector!
- Block level storage is a PITA when dealing with reclaiming used space. It makes using VMware snapshots a nightmare. If I free up the space Ina LUN the SAN should be smart enough to realize the space is no longer used!
- FC for VNXe... can't wait for it! and make block dedupe for VNX arrays to bury this f****g NetApp :)
- Well, they are doing it. But i wish i got the enterprise plus license to play with all the goodies.
- Have easy to use management interfaces [to be fair, not used EMC interfaces yet, mainly referencing Dell, NetApp and IBM]
- VMotion like datastores from active controller to another active controller.. aka a Controller is too busy? and the other one not so much? Instead of only having 2 active / active controllers why not a cluster of controllers? or a head for each drawer of disk to each drawer of disk you get that much more power and performance!
- A consolidated compute and storage in one box. Something like nutanix. The V block is a step in the right direction but I want to see what is beyond this.
- lower their prices. offer models that are just disk and controllers and let VMware do the fancy stuff. Storage tiering...
- Provide for better cross-vendor replication technology. We have to fight with our hosting providers to go between data centers, since no one wants to take ownership of either the data or the process.
- Replication on vm level, might be there from someone. Figure out how to go faster then light, ie active-actice-datacenters with long distance...
- Get Flash SSD to a price competitive level compared with SAS, i.e. replace as Tier 1 Bring storage to allow stretched clusters like V-PLEX down in cost to make them available to the SMB market. Site transparency is the ultimate goal even at the SMB level.
- being able to tell bandwidth requirements for replication without engaging vendors. being able to see pricing for advanced features/expansions without being harassed by sales people
- better integration between vsphere client and emc storage (without having to install plugin & configure SE)
- Have an option for embedding HPC filesystem clients on the array, e.g. a bundle offering a lustre client on the controller, etc.
- Project Lightning looks promising Project Lithium from VMware is also way cool. If these two project had joined forces, imagine the extreme coolness we could achieve?! Veeam's agentless backups are awesome; backup vendors have much to learn from them
- We've only just started on the path with EMC but what I am missing is how Isilon and Atmos play (or don't play) together. Customers need long term, reliable online archiving that is DIRT cheap and never loses a byte. While Atmos has a low $/GB its still not rivaling tape... so we need something that can get there... maybe thats 3TB/4TB SATA thats just down the road?
- multi tier'ed rights management giving vmware admins some access but only a limited amount, and let the vsphere gui interfaces handle this (maybe they do this already)
- implement block level de-dupe to more arrays. Specially in VI environments much of the OS-ses are the same bits and bytes.
- Techbooks - for everything! Especially with EMC's portfolio, it's good to know where things are positioned but it's more interesting to know how things are deployed.
- They are doing great things but the technology is out of reach during tough economic times. Would love to have things like fast/fast cache or even evaluate compellant or other solutions but it just doesn't make sense when you know the funds aren't available.
- Create a sort of Unified storage manager that can create a smaller range of modular vblock look-a-likes. Current vblocks are to big.
- Would very much like dedupe and flash tiering on our primary VM storage, preferably in small-RU form factor. (I know EMC and some others already do this).
- Make upgrades easier without large downtimes to move from storage devices to new storage during upgrades
- Offering Small Storageproducts like an Array with max 5 HDDs, an SSD Cache or enough Cache for the Controller and iSCSI Connectivity for Small SMB's as an Entry Storage Option.
- EMC Celerra - File System shrink..it grows just fine, but no easy way to shrink it when contents have been reduced.
- Not really sure. What I see right now is that the person implementing the storage needs to know about EMC. All seems very personal. If you are a netapp guy, you go with netapp. If you are an HP guy you go with HP. We had an HP guy :) Not sure if there is anything in there you can do something with. Maybe to make sure their are enough EMC guys in the IT business (more training, cheaper or something?) so that we now would have storage that supported VAAI for example :(
- Build a VMWare plugin that allows administration of their storage unit through the vSphere interface.
- The tighter the integration with VMware the better. I want the array and the host to know what the other one is doing and work together around performance tuning and configuration.
- Build a lot of the functionality in. We're just starting and it's very confusing as to what parts do what function and what licensing overlaps with what parts/function we need.
- ease the performance troubleshooting pain by making easier to correlate VM-level statistics with storage-level statistics, and doing it directly on vCenter.
- Inline or post-process dedupe to save space for smaller array clients. Ability to use lower cost ssd as a tier.
- 1.Release the cool granular storage thing which was presented in a tech preview session at VMworld. 2. put it in vplex so we don't have to worry about which volume is biased were.
- Big Iron (EMC, HP, IBM, HDS, NetApp) need to look at the Tintri model. Get rid of all of the old school storage stuff (RAID groups, LUNS, LUN ID's, Zone, Zone Managment, HBA's, etc)
- Allow vCenter plugins for Vmware admins and Storage admin with permission based controls - giving certain users rights to view certain things... I.E you dont want a vmware admin per se - provisioning storage but you may want them to be able to see performance and capacity related things as well.
- From a storage pov im finding it hard t come up with anything on the spot. Software - Hurry up moving other applications into Unisphere e.g. NetWorker, Replication Manager.
- I wish that you'd rethink hw you interact with SRM on the Symmetrix array line. I find setting up SRM to be pretty painful. I'd like to see you do more testing of 3rd party modules ( PP/VE ) before they go GA ( it doesn't play well with host profiles ). I'd like to see you write a decent vApp that can monitor PP/VE ( and yes I've looked at the current solution around that ). I'd like to see you change your licensing model with PP/VE as well to be integrated with the VIC. I'd also like to see you do a *lot* of performance improvements with the VSI integration as it takes a really long time to get any useful information out of it at our size.
- Netapp gives there version of Powerpath away for free, however EMC charges. I think that the ability to utilize both paths would be freaking awesome in vCenter.
- Better/Larger Cacheing Better Built In Performance Monitoring/Reporting Better Built In Capacity Monitoring and Reporting More Cost effective storage Better/Smarter ways to end-of-life equipment...modular upgrades
- Make deploying stretched clusters simpler and easier to do. Array based QoS that understands SOIC.
- Integrate storage and hosts into 1 redundant unit so that adding capacity automatically adds compute, ram, and storage
- I wish they offered more accessible ways of playing with and testing their hardware online or through virtualized storage like Starwind.
- All of my equipment is currently low end, so I'm not sure what EMC has baked into the product line. Proactive notification of storage contention issues due to tie ins to the hypervisor layer would be nice.
- 1) It would be really cool if SSD drives were separated from regular disk on the storage array and ESXi hosts could use them like local RAM. Obviously a 120GB SSD drive would be cheaper than 120GB of RAM on the host, plus it could be host independent! I think the storage controller would need a logical division in processing to facilitate RAM-like speeds from across the wire.... I don't know if this makes a ton of sense, but it would be cool!
- Have VMware-specific storage - like the Tintri offering (which looks awesome but is too expensive for us to buy in to at this time). The range of storage offerings, even from one vendor, is mind-blowing - and each vendor claims "ours is the best/fastest/lowest TCO" etc... where are the reviews/lab tests??
- Something I can afford. (I swear the EMC mgmt learned from the IBM school of "oh, you actually want it to work? That'll be an extra $100K") Yes I jest. Sorta. A little. Well...
- Block-level automatic data tiering (a la Compellent Data Progression). Pay-as-you-consume storage with additional capacity/performance disks/shelves on the floor ready to go.
- EMC holding EMCWorld in Asia Pacific region!!! Like SAP or VMWare holding their events multiple locations in the world.
- We need cheaper and more scalable storage. I want to use something more like OpenStack Swift for my VM storage. I want thousands of desktop-class spindles spread across many "nodes", with data replicated across multiple nodes, instead of hundreds of expensive enterprise spindles, with too much expensive enterprise redundancy connecting them. But unlike OpenStack, I want to present a standard block-level transport (likely iSCSI considering each "node" would be a cheap server) instead of an HTTP-based object-level API.
- I wish EMC had a truly unified block/file storage platform similar to NetApp. And don't say VNX.
- Tie it all together so I can see which VMs are hitting datastores hardest and how that is handled by the underlying storage
- In our case if HP could up their game with the EVA's and start providing better intergration with vSphere i would be a happy man. They are capable, if simple units, but the management leaves a little to be desired.
- Want a secure way to give rights to use vaai in a single vmax, multi tenant model, but still maintain per client security
- EMC - Please consider Dedupe on VNX. I know there is Compression but Dedupe would be nice both in a techincal persepective and in an "internal(to us) marketing" perspective. EMC- Isilon. Make the CIFS have their own IP-stack so that we can use it in our environment. We're a service Provider with a lot of customers with the same IP-subnets.
- Make snapshots as easy as HP command view. Not need multiple systems and software to accomplish simple tasks a EVA can
- Gave away free hardware to consultants :) Some online training with remote kit would be useful for study.
- atmos needs to go away. that functionality needs to be integrated into the vnx line as an add on.
- Block level DeDupe in my VNX5300! Some VMs dont need performance, so why do I care if I take a performance hit for the deduped LUNs?
- Better included integration with virtual environments so I can auto-deploy new VMs without writing my own scripts to determine what datastore to place new VMs on. I know the integration is there but I still have to convince my management to buy it! :)
- Make it possible to easily shrink VMDKs from within VMware. This may be possible now. I only get to touch very expensive hardware at customer sites and I'm usually just performing initial configurations.
- We would like to use stretched clusters with VPLEX. Today we have the feeling, we just loose to many features by virtualizing the back-end, which makes the decision difficult. A VMAXE is a really expensive array with lots of great features, we would still like to use, when it is hooked up to a VPLEX
- Make more simulators available to learn the technologies. We currently use one storage vendor, EMC. If I ever move to another job or start consulting, i'll run into different environments that have storage other than EMC. Need a hands-on way to learn without paying 5K or more on a class.
- Better VM storage tearing capabilities. Improved & seamless backup solutions on the storage-arrays.
- Deliver a constant performance :) And to stop charging over the top prices for disks that are done by Seagate all the time..Focus on the price for extra functionality given by software rather than on commodity hardware like disks.. The possibility of adding memory to the storage array, like adding it to a server and voilá..instantaneous performance boost with mega caching effect.
- Improve features like expansion of a mirrored primary and secondary copy (mirrorview) without deleting the mirror, possibility to create wil on pools as also CPL.
- Should not be so expensive to get the speed. FLASH is insanely expensive on EMC. Hence we went with WhipTail for our new UCS cluster.
- Better management tools and end to end integration. PP/VE specifically sucks in this reguard today (Chad has heard my rant on this). Better cross vendor qa of plugin modules and add one for compatibility
- Dunno. I guess management tools could be made better, more performance, more suited to ESX. I think it would be cool if these vendors (I'm talking about HP in particular) would make their support a lot better. Ever tried to find THE info you're looking for on the HP website? It's a PITA. And then we're not even talking about getting a replacement disk...
- I spend so much time trying to keep up with new features that I dont have time to think of them on my own. :)
- Simplify storage snapshots + replication and clustering and let vCenter Server handle or atleast display information regarding storage, so that one tool (vSphere Client) gives me all the info I need on the entire virtualization platform.
- Better integration with application understanding for snapshotting. Intelligent snapshotting, much like the auto tiering of Dell Compellent placing snapshots on cheaper disk!
- a) Be truly inter-operable between vendors (i.e. 3PAR to EMC remote replication, etc. b) Make features like replication, deduplication, multipathing, and capacity not a license restriction.
- Fix UIM. It downright sucks. ( I hear v3 is better ) Get Emc to buy apg watch4net or similar so we can do true capacity and performance managment across the stack.
- Cool feature like large amount of flash cache in a storage virtualisation appliance is a nice thing. We're doing that already with redundant clusters of SANRAD XL's.
- Have the auto-tiering policy being directly driven by application rather than manually setting these up based on pattern that we think we understand and possibly pre-stage data before the batch or jab kicks in
- Use standard storage for backups leveraging snapshots instead of specialty products like Data Domain/Avamar) Like snapvault for NetApp. Make ECC easier to use.
- Implement storage hypervisor idiom : Abstract layer between hosts and storage arrays making storage migrations (switching arrays) seamless and painless.
- I would like to see EMC be less arrogant/overpriced with their solutions. Completely ripped out multiple DMX/Celerra and replaced with HDS and could not be happier. Our enviro=EMC free shop. But I do like your blog so thanks for that.
- Simplify and mitigate the requirements for metro/stretch clustering of storage systems to come close to VMware requirements around HA/DRS/vMotion
- 1. An easy method to increase the LUN's size which are currently being presented by the VPLEX 2. An easy method to delete LUN's being presented by VPLEX.
- Better insight into auto tiering of storage. Which VMs caused the blocks to be tired higher. better reporting
- better documentation, better KB's; especially on fringe products like, say, plugins for backing up Exchange mailboxes using array methods
- Proactive performance tiering. Looking at long term patterns of usage and auto-tiering volumes up or down based on the pattern. For example finacial reporting at the end of the month.
- Stop charging 10x over market prices for spinning disks. And don't get me started on the EFD costs, either.
- EMC needs to get the hole-punching that NetApp just released. Oh wait, they need to get block level dedupe first...
- Since the major players can now "see" the virtual machine, it would be nice for them to report actual IO the vm is using on the array.
- Less restrictions on the entry level storage arrays, or at least a better commitment that they won't be left behind when useful new features are introduced in other members of the product family (AX vs CX series, VNXe 3100 vs 3300, etc)
- I always love to see more unity between management consoles for different features. Even with recent improvements, it still feels like there's so many "panes of glass" to manage.
- HP Storage and the F%&#ing VAAI that they have been promising for the EVA (or P4000/P6000 as they are now called). I wish we had some EMC storage :( HP storage can suck my .....
- The new profile based storage is a step in the right direction. Being able to inteligently provision vm's using powershell cmdlets. Anything that makes this process simpler would save man hours.
- Consistent release strategy. Or at least a clearer message on what works with what and which features. It's sometimes a struggle to keep up with that between constant release cycles of things like Unisphere, DART, vSphere, etc. Add in other tools such as SRM and we don't want to accidentally tell a customer something will work when it won't.
- Ability to easily add vendor agnostic storage behind the controller as a tier. ie; NFS sata from another vendor.
- Push VMware to make NFS multipathing more straightforward (no more single TCP sessions and having to setup virtual IP addresses to balance load).
- Better integration with VMware concepts. This is a mini Compellent rant, as it looks like we're buying 30-40TB of Compellent next year, and I have to throw out my wonderful VMware templates and clone VMs with Compellent's Replay technology instead. I'd just gotten used to how everything worked with NetApp-backed NFS datastores and the switch to Compellent and iSCSI is a bit annoying. No deduplication, really? :(
- Based replication technologies on storage pools to map out to SDRS cluster so change can be tracked at pool level to reduce replication time and data changes as currently there is only based based change tracking or vm based so SDRS is no benefit for DR requirement other than initial placement.
- I wish they would make test/lab equipment available for extended use. I know that's a lot to ask, but if we're testing our apps on it in the lab, there's a better chance we'll purchase that storage for production.
- Stop gouging for Performance and Reporting tools over and above the base software. Licensing even low level aggregation tools is so hard to justify to management when they point at the barely adequate vCenter or Analyzer graphs and say - 'There you have reporting!'
- Free, or ultra low cost price points for development platforms that focus on storage heavy deployments. I think NetApp is ahead of the game here, wish EMC was more like NetApp in that way.
- Let's get something similar to Solidfire, an all SSD storage solution with enough IOPS to make the highest end spindle systems looks like they are standing still.
- More efficient storage configuration interface - other than EMC vendors can do a lot of things more user friendly :)
- More literature targeting SMB's. EMC is still viewed as expensive and not fit for SMB environment.
- - Better VMware integration ... it's coming but not fast enough. - Even more offloading ... doing things in software is plain stupid when you can do it directly in hardware (more VAAI, using features of new processors ...) - Real tools to manage the whole
- Fully integrate lun/storage management into the vcenter client. I don't want to use extra utilities to simply provision a LUN.
- We're just strapped for time, figuring out how to fully utilize the product is the last thing on the list. The product gets installed and configured for the environment at that time, but six months later the environment's changed and the product could use a tune up.
- Better integration with vSphere, easier replication of specific VMs, easier provisioning of datastores
- provides the highest I/O read and writes for preventing I/O contention when implementing large scale VDI project
- I find that every vendor needs to follow a standard set of protocols. If I create a script for one storage solution, like I can talk to a netapp with RSH in a script to do snapshots, but then if I use an equallogic I have to setup python scripts to almost do the same thing. Except I can't rename a snapshot on a EQ but can on a netapp.
- Remote replication of Iscsi from within the VNXe not just NFS (not that the sales guys tell you that at the time).
- Stop riding the marketing and buzzword bandwagons and getting generally more chad'astic. Like in doing more multi-vendor initiatives and stuff.
- failover to another storage in case of failure and integrate that with vcenter for failover of vms to be automatic in case of storage failure
- Make storage easy. I'm not interested in figuring out RAID Sets etc. Just do it for me and give me space!
- Sell storage software (isilon / VNX stack) that can be used on older DAS arrays to provide a low tier / target replication tier.
- Make auto-tiering more user friendly. Same RAID type makes growing/changing/configuration hard. I like controlling RAID type so I can maximize performance/capacity as needed. Just wish I could do RAID6 for the SATA stuff, and maybe RAID10 for the fast stuff. Maximize both ends. I know some that want to just add disk and let your magic voodoo do it's stuff.
- Keep up with VMware products and features in time (Networker integration with vSphere, Vmware SRM 5 MirrorView adapter)
- Storage, CPU, Memory in one. Storage and compute resources in a single easy to deploy device that integrates seamlessly with vSphere that isn't VCE vBlock.
- I wish emc should start providing cloud virtual machines on flash drives As the world is moving towards tablets emc should come up with android and iPad apps for provisioning storage,DR etc
- Build VM aware storage and backups into array / vCenter plugins - make them ready for primetime, and ditch Java (more web / vcenter friendly)
- Provide online Labs for Partners to test/study different storage technologies like the labs at VMworld and at the VMUGs.
- EMC are always providing us with Lab time and equipment to help us further develop the IP projects we are working on. Without EMC's assistance and partnership with VMware our jobs would be a lot harder. Thank you for your support to the VMware CoE
- Integrate array performance statistics into vCenter. Also integrate more array management pieces into vCenter. Ideally, almost all array admin functions as it relates to VMware could be performed in vCenter.
- cheaper SME storage that isn't crap. The iomega line is a great start. The refresh has been stellar. More hooks into RSA. enVision and Archer are good starts. vShield Edge needs to plug directly into RSA DLP for management.
- Really cool easy to use storage interface that would gives you views from the storage, from the host and from the VM's without having to recreate the world
- implement federation in lower end arrays , whch helps in long distance vmotion and disaster avoidance
- Compy with VAAI and VASA, It's crayzie that some storage vendors have VAAI on the 2012-2013 roadmap!
- VPLEX is an amaizing, but I worry about manual switch to secondary site if primary site goes down. Management integration with vSphere Client. I want to use only one instrument for any type of works with infrastructure - VM/ESX configuration or storage performance tuning. All-in one solution for manage virtual infrastructure - my dream.
- I've commented before that my experience with EMC was very negative not because of the technology, which is great, but because of the salespeople abusing us. I wish EMC would get sales people that aren't Ferengi, and actually wanted to work as part of a team with the customer, instead of being yet another way my life in IT is complex. You guys are doing great things with the tech (and I do wish I had one of those VNX(e) boxes!), but I really think the war is being lost by your front line sales guys and their scorched earth tactics.
- Better whole picture performance troubleshooting. From VM-DISK. What prosphere looks like it will do.
- Make true mirroring from site to site and SRM an easier task. The support environment in EMC is also brutal.
- Make larger amounts of reliable, good performing, easy to manage storage available for less money.
- Simplify Storage Management. I have the impression IT from the Management point of view has not developed since the last 25 years.
- Can't think of any now. I'm sure we're not using a lot of new features that are provided to us.
- Maybe I just am missing it, but offer a full solution for backups(to disk) explain retention/drive swap, and DR appliance.
- be able to know how much storage i am using in both capacity and performance, and predict for the future how much will i need
- Provide storage arrays as virtual appliances for testing and educational use, like you have already done for some.
- array simulators or virtualization projects for lab and testing use. Something I can use at my lab at home that is pretty good (9/10ths) (not production, but not so hamstrung and horrible that I would want to scratch my eyes out) Oh and make it affordable.
- In regards to Netapp (could also apply to EMC maybe) be able to snapvault and snapmirror a primary volume to the same secondary volume, they share the same data blocks. Currently have to vault & mirror 1 primary volume to 2 secondaries and use twice as much space, have 2 copies of the primary data and can't dedupe as in separate volumes. Makes DR and backup\Restore very space intensive, we have to chose one or the other.
- Automatically recreate virtual memory disks (page/swap) at a DR site so that GBs of useless data doesn't need to replicated.
- Simplify. We had two EMC techs involved with the last SAN we added and it took over a week of scheduling to get it finished. The EqualLogic SAN on the other hand was installed in a few hours. The administrative tools and concepts are built for dedicated SAN administrators and considerably hinder people wearing many hats.
- PowerShell ALL THE THINGS! Seriously. We use powershell as the primary reporting mechanism for the virtual environment, but I've found EMC's cmdlets to be frustratingly lackluster.
- storage is so difficult to learn. more virtual appliances to learn the basics with more functionality would be fanatastic. the more admins that could use a test system like that the more that would be comfortable. as it is it is very difficult to begin to get storage experience.
- Get array DRAM cache memory (Non FAST) to level that supports todays requirements, few gigs is nothing.
- Maybe lack of knowledge on my part, but a kind of `service profile`for best practice configurations and the vendor`s array ie this will be a vsphere-SRM- replicated lun.... too many variations, i know
- Create more "free" virtual storage appliances so we can have a better look and feel experience, EMC is doing it but nog all other vendors...
- I wish they would stop throwing FUD around. EMC does a pretty good job of keeping that to a minimum, but some of our other vendors are a constant struggle. I wish EMC and our other vendors would allow us to do capacity on demand in the storage space, so that we could get all of the gear in advance, then true up as we start using it. This would make bursting SO much easier than waiting for A) a project to charge against and B) waiting for EMC and other vendors to deliver the gear
- Give my access to project lightning demo unit;-) Integration of vm performance data in the VNX analyzer.
- Better Performance Monitoring on Midrange (VNX/CX) - The build in functionality is really not sufficient
- reduce costs :) seriously : agree on the common storage standards for all the major functions so I dont have to learn new management "paradigm" every time I switch the storage vendor
- EMC is doing a good job integrating with vCenter. I would like to see them continue on that path...deeper integration.
- #NAME?
- EMC to have easy to read quotes/invoices broken out by line item and true cost of each component. We have had line items that read 146 units at $7.43 a piece?!?!
- provide better/easier to access test platforms like the ubernas appliance. Let us test features in a cheap/free environment before using them on systems that cost big bucks
- Make it more easy to reclaim unused storage (thin provisioning). Ease of use in general, also multi tenancy ( cloud enablement)
- Make sure all commands lines are powershell. Understand that the price points for small offices needs to come down.
- Would love some vcenter built in EMC help, such as "Hey, your using NFS on VNX, best practice is X number of VMs on this datastore." etc. etc. Little notes go a long way!
- VMware specific Storage online trainings. These trainings will be focused on using SRM and EMC. Storage DRS, Storage best practises. Configuring Storage for performance
- More info into vCenter.....If i could have full storage management from within vCenter i would be a happy man
- stop nickel and diming for basic functionality. go through whats included in the 'basic' package, and include the stuff that admin's actually need to stay sane instead of making it all add on packs. best practice analyzers.
- Integrate block storage in VMware so that vmfs expansion or shrink is a single hw and software event. Replace zoning with something easy and automated.....like when I Lin mask, auto zone based on that.....
- Do a API for the customer, that we can use as a whitelabel solution insted of developing it our self.
- Get the OS off the drives and on to the controllers ( like Equallogic). One OS across the lines with clear and easy upgrade paths.
- Virtualized network ad storage connections on the arrays. Can assist with being able to dynamically migrate connections to replace hardware components.
- Better visibility into the underlying performance metrics without having to pay for additional bolt-on license/upgrades & even then they are crap! (I'm looking at you, HP)
- Make managing storage easier as it seems to be hard to rightly manage a big System like 3PAR (longer story...)
- Make an Iomega-class device that can replicate to another Iomega device across the public internet.
- Most of the cool features missing in storage (i.e. storage drs) are now addressed by VMware - although there is still a disconnect between a provisioned VM and the performance/behaviour reported from the storage device to isolate VM related problems.
- Ditch the SSD drives for solid state storage onboard (PCIe) and rewrite file systems to efficiently handle meta data updates and not thrash the solid state storage
- Simply management (aka VNXe) - currently too many tools, interfaces to get things up and running, look and feel different. Also, Unified Wire - a single wire which can deliver FCoE and NFS. Currently too many cards each doing its own thing.
- A single performance tool where we can view all perf asPects from the physical. Ladd down to the spindle.
- We have all the building blocks for VBlock but do not have it. This is the second company I have been at where we could be VBlock certified but are not.
- Build some kind of storage virtualisation appliance that persents the arrays to the servers (hardware box). Move the storage in into the esx servers like you showed with isillion! I think this could be the future and would simplify things.
- Make it easier to provide reporting on storage usage and allocation not just from a blackspace perspective, but also I/O, cache, etc.
- Better price like VNXE but with higher performance. VNXE use only 1 gig of ram for write cache, need more!
- I've gotta say, eliminating RAID is pretty big in my book. Simply presenting LUNs and not worrying about the RAID type, tier, etc. makes it a lot easier. One type of storage that can meet all of our needs makes it far easier when talking to our customers, since it takes significant variables off the table.
- More plugins to vCenter, allow us to do more from central console (create/resize/snapshot/replicate/etc.).
- better storage api integration, Can't really comment on the EMC side as we have HP storage and in my personal lab I use the cellera virtual storage appliance
- * Block level deduplication! * A more cool performance tool integrated with EMC Unisphere. The graphs seem rather "primitive". I personally like SAN Headquarters from DELL Equallogic! * Decide between HTML and Java Applications for the unisphere management tool, the combination of both is a PITA! * Get rid of some logical objects: Storage Pools, Storage Pools for file, raid groups, metavolumes, LUNs, devices, filesystems... Sometimes it is overwhelming. It looks like EMC haven't got enough time to mix the features from Clarion and Celerra... Thanks to god, the VSI plugin helps a lot! * To Know the TCP connections currently established to a NFS export. Would be great to know which clients use Jumbo Frames from the Unisphere Management tool.
- Isilon with VAAI. Storage vmotion is cool. Upgrading my arrays by adding and removing nodes and not having to storage vmotion thousands of VMs is so much cooler. Cloning on isilon is slow without VAAI. Backups could always be easier.
- Keep up the good work (storage agnostic). This industry is moving fast and the ability for storage vendors to keep up and integrated will be paramount to success.
- Release out dated information about 1st gen and 2nd gen families to build and restore such units from scratch. Like the utility partition upload over serial.
- vPlex is so interesting. DELL Compelent approach for Software and Equalogic price so cheaper. Is more "hard" to justify EMC with my boss.
- MORE WHITE PAPERS. Reference archetectures are awsome to base decisions off of. Thanks for the good work!
- Move to a service-centric model where the storage is but a component used to build a service. With the inteligence built in arrays these days, there shouldn't need to be a big need to design a service around storage if the storage does it's job. When there are posisble capacity issues (let it be storage capacity, IO, DR/Availabilty, or Power/Cooling), the information should be relayed to those that can deal with it in time to resolve any possible reduction in service delivery/availability. From a design perspective, historical information should be easy to get as to make an inform decision. Use Open Standards to manage all arrays Vendor neutral storage API's for all virtualization platform Availability of open source or free tools to manage arrays which work with all vendors (Now I realise that not all arrays support the same features, but that shouldn't be a roadblock to having something that can dynamicaly change based on a set of features that a array can provide.)
- Better customer service. Meeting SLA for replacing failed hardware. Documenting design flaws (ns120 thin provision not fault tolerant).
- Quit confusing my poor customers. All these propritary names for the same damn thing have made the sales cycle for new SANs outrageous and no one knows what they are buying. A simpler way to explain real usable storage depending on features in use (I see people not understanding RAID and snapshot overhead).
- better visibility into performance issues (low iops causing issues, need more spindles, etc) and possible recommendations on remedies would be useful
- Offer a self-replicating SAN-less Google style environment (see Nutanix) converging computing and storage
- Remove the concept of LUNs altogether. Let me just drop VM's onto an array or combination of arrays and it can handle storage vmotion or migrating the bits around. It's a pain to have to care about what goes where and migrate/expand contantly.
- Revisit RDM vs VMFS with Oracle; test most appropriate linux elevator (noop? cfq? deadline?) with VMware and VMware with Oracle DBs? Compare CPU overhead of VMFS3 (supposedly 8% higher than RDM) and VMFS5 (??) Expand FAST Cache drive limits for NS480
- Stop selling VNX file and focus on Isilon as your file strategy, ie isilon integration with vmware, smaller onefs block size to deal with small files and small IOs.
- I'd like a better integration and design of EMC plugin with vCenter. I am generally disappointed by the EMC management tools and especially the integration with VMware. While the partnership is developing technology and the backend is good.
- Look at what some of the smaller niche vendors are doing and consider expanding their lines to include some of these features. Ie deduplication in cache (tintri), array of redundant nodes (ibm xiv etc).. auto alignment of partitions etc etc
- There's a ton of cool stuff that we aren't even touching, so if anything... I wish we could get our hands on more stuff!
- Automatically identify files and data that do not really need to be replicated for SRM. (Swap files for instance)
- Midsize storage system which could be split into two parts and installed into different datacenters.
- Cheaper SSDs :-) Reduce power usage in the Datacentre. More deeper integration with vSphere. More reference designs of hardware/software.
- Unsure been looking at all the new features of 5 a d we are not even close to implementing it at my employer. Did really enjoy a vBrick demo at a local VMUG I'm sure that will catch on.
- More cool virtual appliances, using Celerra Uber in lab now, best marketing you could have. Now being able to simulate v-plex would be awesome :)
- to be perfectly honest I would like EMC to produce more reliable code for Celerra or VNX - File. you guys have a very robust hardware. your Clariion is probably, and based on my personal experience with other vendors, one of the best mid-range arrays but your DART is a pain in the A**. You should slow down your run for new features and take longer quality checking before releasing new versions. That would be very cool from EMC.
- The VSI plugin is weak. I really wish VMware knew how much pool was used on the back end, so I wouldn't have to build giant, oversized datastores to get around the fact that it doesn't know.
- Replication and DR for SMB, where if I could have VNXe that replicates 3-5 times a day (snapshot-based) to EMC PaaS Cloud based service and if disaster happens, I can just start all replicated VMs in minutes, reconfigure the DNS to point to new public IP(s) and continue the service (file shares, Exchange, SharePoint, etc)
- Primary storage block dedup for vnx, better preformance analyzation for flash/cache accelerator in vnx
- support storage providers for CX class san, and have a easy download page/doco to use storage providers in vSphere
- Drop costs! :-) Seriously, with the healthy competition in the storage industry, storage vendors are staying on the cutting edge, providing cool features to the customer. Well done.
- Would like to see a good-performing small iSCSI array, LUN size not limited to 2TB, as a possible backup-to-disk solution. And have a removable component for offsite backups.
- EMC and other vendor are already doing cool things, but good use of flash drives with FAST is a good area. I don't want to have to sweat over swapping.
- - more cache in storage array's - better/fast storage controllers for handling high load and redundancy
- SRA enabler for MirrorView (I know it coming but still ..) Storage Profile support when using FCoE (maybe this is more Vmware problem)
- Actually try to solve our problems instead of saying "buy a vblock and this will all go away". Put foot down with our storage team about sharing FA ports and doing things the right way.
- Personally I would like to deepdive into mechanism in the array so I can design it correctly. Still the a lots of its and buts. for example FAST VP. in emc I want auto balance for FAST VP now so we can add disks to the pool in increments of 5 even when its populated.
- More granular and deeper integration for storage with all hypervisors - I'm aware of VMwares future vVolume wrong with vendors - but similar needed for hyper-v as well
- We currently use VNXe for our VM storage. One thing that would be nice is just to add SNMP polling ontot he VNXe arrays. Something so simple should be included in that for monitoring...shocked me that it wasnt.
- We use the IBM Storwize V7000, and the biggest feature missing from this (IMO) is inline deduplication.
- I wish storage wasn't a roadblock to virtual desktops. The sheer cost of storage alone to puts things like virtual desktops just out of reach, and then the licensing pushes further away. I wish EMC would leverage the existing storage of servers on hand as storage pools for virtual desktops.
- Streamline their basic mirrorview replication so that it does not require so much manual intervention.
- Make storage into a a true internal cloud provided service. I want a single namespace connecting to all my data wherever it is and get the storage to manage replication / availability
- Team with BMC more for better instruments integration. Introduce standalone mini NAS (not VNX-based) in a 2U rackmount format etc, expandable via iSCSI or even SAS BE.
- Break the barrierer between SAN and NAS. I want to allocate space and not have to worry whether it's SAN or NAS.
- More integration from the array into vCenter. Extend view like storage to show natively things like real storage use on a DS, even with dedupe and thin provisioning
- Giving us a plugin for the vCenter Server, so that we can see the performance from the storage view would be fine and make troubleshooting easier. Giving the availibilty of the sized LUN back to the Host and naking automaticly storage classes maybe gold silver bronze
- Migrate automagically from current arrays to legacy arrays when resources require - I mean that the underlying management should be able to "speak" HP so that I can continue to use EVA storage when I have moved on to VMAX or VNX
- Help the storage and VMware customers have an appropriate dialogue about configurations. Abstraction of device configuration makes SLA determination difficult.
- Simplify storage. Provide a mechanism, similar to netflow for networks, to visualize and break down my storage IO so I can see what and where my traffic is. Right now responsiveness of storage is kind of a black hole. You can see it at one end or the other, but what happens between those two points is frequently unknown. There has to be better visibility into that.
- simplify management, I'm used to equallogic and it's easy, wish other vendors would work the same.
- Help the small shops out with prices... we want to use the cool stuff, but don't have the cash to back it.
- Bundle internet bandwidth with their storage offerings by partnering with core service providers hence making it easy to take data offsite.
- Since I haven't reached the limits of what storage vendors are doing already, I'm unsure what I would need beyond that.
- I think EMC does a fantastic job on showing how its products integrate into VMware. I think other storage vendors could learn a lot from that (which might not be good for EMC)
- Can't say as I haven't had the chance to play with vcenter/VNX integration yet. Maybe I'll have a wishlist after that.
- enhance the EMC vCenter plugins to display the tiering information (% SSD/FC/SATA) a VM resides on.
- More plugins/integrations to vCenter servers so all single glass of management and provide better performance and throughputs and more granular troubleshooting details deep down to the weed but from vCenter instead of Ionix Control Center
- Thin provisioning reclamation for both VMFS LUNS and RAW luns. For SOIC to be aware of other RAW iSCSI mappings.
- Publish more specifics about VMware type options and recommendations, a more deep dive of information regarding specific options that are available and their impact. I want to know what NEGATIVELY affects my array, as well as positive.
- ability to link several arrays together as a single unit (ie get past the 2PB capacity limit in a CX4 series)
- Increase the lower end storage options, not everyone can afford enterprise storage yet most people have the same storage issues.
- Would be nice to see some EMC provided equipment in local community college and technical schools.
- Come in and replace my Hitachi. I wish the Hitachi had the same level of integration with VMware that EMC does. I'd love to be able to see more deeply into the storage aspect inside my VI client on the Hitachi USP-V side or even the AMS side.
- Simple management Performance measure features for storage and VM Efficient and cost effective replication over WAN
- make it easier for the user to understand performance bottlenecks and give clear procedure recommendations. Referencing expert level documentation kind of sucks.
- Smaller appliances to compete with the startups (Tintri, Nimble) for a simple, cheap black box storage
- TIghter lock step in terms of release with VMware et al (see MV SRA as an example). Mini Isilon would be cool. And a virtual Vplex :-)
- Quit screwing around with small DRAM caches. I have 128GB on a vSphere server. Why don't I have 128GB (or more) on a VNX or FAS? DRAM is cheap relative to the overall expense of a storage array. SSD caching is great, but since first write hits disk, not SSD (FASTCache), it's appeal is limited in write-heavy environments. A dedicated 100GB+ DRAM battery-backed write cache could have dramatic improvements on performance and cost of storage arrays, allowing us to "hide" the write performance of NL-SAS or SATA behind huge DRAM caches.
- I'am very frustrated when implemented EMC VNX doesn't work well in the area of management (poor Unisphere). When using Unisphere we often get different results using different computers! EMC please design it better. It is VERY frustrating to our coustomers and very often IBM wins because of that (even without VAAI and VASA).
- I can't complain too much about EMC's products, what can I say.. we're spoiled! I can complain a lot about other vendors products ;) See below for my wish list on EMC (issues outside of tech/products) in particular.
- Better reporting tools with costing, ie put in the cost per GB and it works out the cost per a customer
- I think storage vendors are doing some amazing things already that are much further along than the other infrastructure technologies related to virtualization.
- Tighter integration as to regard of the vplex and vCenter. Moving away from DRS/HA per vm configuration to a underlying storage concideration as to how act upon different scenarios (DR or routine)
- Certain vendors should run more "hands on labs" events around the world to get familiar with their storage and attract more buyers. I would love more events in Europe.
- VM-based replication (with de-dupe) between datacenters (like Zerto). Also, and this is huge....BACKUPS! We have EMC's Networker and quite frankly, it's terrible. With all the cool features VMWare has and all the cool features EMC's storage has, why can't they work together? I won't go on and on but it's been one headache after another.
- Analysis of a VM environment and auto LUN sizing and tiering based on available metrics. Not saying I don't want to do that stuff myself (love design as much as implementation), but doing that at the moment would buy me the time to design and implement some of the cooler VMware technologies on top of storage without thinking that I had neglected one of the more important layers of my solution. Pie in the sky I know...
- Make better management interfaces. Unisphere still is a bit much. Also, better real-time statistics (other than running the analyser and then looking at the info you actually want to see.
- Drop the price of FCOE based SAN by 75% Easily managed fully meshed or 1 to many geodistributed LUN replication via DCB. Cost scale to support Hybrid RAID Cloud Architecture. Publish a Reference Hybrid RAID Cloud Architecture design
- -full integration (storage capabilities negotioaitons, sla, sdrs, tiering, .... all in one block that would be negitiated trought standard protocol and supported by virtualizations layer and storage vendors) -ease diaster recovery: disk resiliancy, replic
- I wish them to lower prices, that would be really cool. And vendor-agnostic array replication would be cool too.
- Finer grain control of auto tiering, priorities to work loads based upon time of day. Also, need to be able to exclude objects from auto tiering based upon time of day.
- I think everyone in the storage arena need to make it easier to determine performance of storage arrays. I'm not a storage guru, but I know that IOPS is the basic storage metric, but in all of the performance monitors in VMware and most SANs that I've seen they show throughput or response time. I know they are useful numbers but a majority of customers just want to know basic stoplight levels of performance green/yellow/red. Make it easier for me to show that it's not the underlying storage which is causing application performance please.
- Would love to have the Storage Space reclamation work again w/o any impact. Also would love to see the costs of EFD's come down that can help customers do more and more Storage Tiering for VDI work loads.
- Provide public labs that end users can sign up to demo equipment without having to go to a road show.
- Start making servers :) Again! Finally get Dedupe in Midrange Streamline Unisphere even more
- further integrate with VMware VC to allow storage provisioning directly from interface. Continued development of failback for SRM features on array level.
- Make it easier to find performance problems. Navisphere is a great tool and exposes a lot of information, but when you're in a crisis, it would be nice to see a screen with a big red arrow pointing to the issue that's causing the problem instead of having to slowly track it down. We're looking at vcenter operations, vkernel, etc to try to get better information in a faster amount of time.
- Create very cheap models for people like me to learn with. So we don't use geriatric things that if were human would have gone through dignitas a long time ago.
- I would like to see more best practice whitepapers and information for people using some of the smaller shops. Lots of papers even from VMware, show examples of maybe how to run a 40,000 mailbox cluster using VMware and uses over a quater of a million dollars worth over storage (not an exact example). How about a 2-4 exchange server cluster running 400-2000 mailboxes. What kind of solution works well in this instance?
- Pick of the pace on project lighting. Fix the Vcenter plugins, they are horribly complex and buggy.
- Throw some tools into the Power GUI Power Packs site. http://powergui.org/index.jspa More free tools for the shops that cant get a budget for cool $tuff. Like us Local government geeks. Let Geeks with VMware licenses play in the EMC-VMware Demo/LAB. It is hard to get a place to learn and test ideas and knowledge. You could see where we get into trouble and use that info to fix future UI's.
- Create one standard storage API to interconnect all the vendors and make a single point of management between all vendors
- Make their GUI's super quick and less dependent of outdated JRE builds. Offer sub-LUN tiering in their SMB/Home-lab gear (PX4/6).
- The software licensing (CIFS, NFS, ETC) is a source of frustration for me. I would like to buy the software capabilities once and refresh the hardware as necessary. We lease everything and I feel like i've paid for our CIFS and NFS license 3 times over.
- provide, publish, document and maintain APIs for provisioning and configuration. provide integration e.g. in VMware orchestrator or Powershell (I prefer the first :-) )
- Raid-DP like capabilities with read /write SSD caches that auto adjust based on block size IO operations
- Make backup easier but directly exposing Snapshots to a linear backup process (2Tape or 2File). Deduplication in the backup process.
- Common replication protocol Simplify Licencing, why pay extra for replication licences when you could pay VMware for storage agnostic replication. Help push MS into supporting Exchange running on VMDKs on NFS Datastores.
- Give more visibility and control into what EMC FAST is doing. Make storage array aware of tasks like sVmotion, TP Reclaim, so they will apply QoS for these tasks to not affect other workloads
- Make the cool stuff cheaper. We feel like we're rebuying the same software over and over on new arrays.
- Remove hard drive bay design for PCIe based back end, and SAS protocal. Increase storage shelf density Increased integration of FCoE
- Promote the abolition of siloed administration of storage - particularly when the environment in questions is over 90% virtualized. I actually think EMC is promoting this, but it still seems like the local sales and service team is contributing to "silos". Target the infrastructure teams as a whole, not just "the storage guys."
- Better backup solutions for application consistent vmware backups. Easily fix vmware alignment challenges.
- Much much much better UI. Storge is moving from the 100% specialised uber-storage-geek to the server guy now, so we need to make it intuitive.
- I wish EMC would make a low-end, affordable SAN. We have lots of small affiliate sites that really want to run on single ESXi hosts to save costs but of course we want at least 2 ESXi hosts with shared storage. IOMega product line fits the bill here in terms of performance/price we need, but we are a large enterprise (multiple V-Max) and the storage team wants to use standard management tools which apparently they can't do with IOMega. So we need an EMC SAN to solve this problem.
- I wish EMC would bring tiers and flash cache on VNXe. Would also be great to get hands on with the Analyzer helper.
- Make replication technologies more granular and mesh better with vmware technologies. For instance, since consistency groups are based around LUNs, it is limiting us to a very small amount of vm's per lun (since we want to use SRM and have to failover each application on an individual basis). Which means lots of small luns, which means lots of wasted overhead space, for vmware snapshots etc., and that many luns in general slows vmware down (ie. rescans). So we'll have 100 200GB luns vs a handful of 2TB luns. It really limits SRM's usefulness or VMware's usefulness in replication.
- I wish EMC would spread the word about Storage plugins and VAAI a bit more. I had to tell our Storage guys aobut it and they're EMC certified!
- Multi-vendor inter-operable standard for LUN replication over IP networks with throttling. It's ridiculous that this doesn't exist in the proprietary world but it can be achieved in the open source world (drbd) on Linux.
- We talk about this at EBC's. If you would like to attend one and hear what we have to say consider yourself invite. Talk to Jerry Eudey @ emc and he can put you in touch with us.
- Open up more documentations on how system handle data/block internally. Open source file system, like ZFS, one can easily understand how system works in any given situation by various mail list, forum, and source code. However for EMC, Netapp, or other storage vendor it is like a black box and customers at mercy of trusting sale rep.
- Provide more root cause analysis ability within the core monitoring platforms supporting storage. All too often there's a great amount of data but pinpointing root causes tend to still require the array vendor to be involved.
- Slow down! New features come at such a pace that its difficult to implement / investigate them all. Seriously though - its all good. More options is always better for me as a customer. I think EMC are doing great things around making things "just work" without having to go through massive amounts of hoops and configurations steps. Making further strides on this path is key. Make VNXe-like ease of use enhancements to the complete product line. Make storage admin guys and server guys redundant. We want to be virtualisation guys!
- I want you to make a storage that an individual can afford it for learning/testing curve. It should be packed with features of the real storage but with cheap parts. Terabytes of storage and high speeds are not that important in a testing/learning environment. There are many things that many of us in virtualization would like to try in our own personal lab but storage prices are to high for some products that are too advanced to be used as a playing ground. Cheap housing with few sata disks 2.5" or 3.5" with some average controller and a four nic ports would be enough. But most important software support to really try most of the features!
- The documentation, blogs, articles around vmware and its uses should be the gold standard for other vendors in IT.
- More details on FAST locality, please give more tools to the customers. like Analyzer helper, and all the other tools you talked about in a webcast. Lets know what the SP´s are doing when they run at 80 %
- Make a more intuitive interface. Using Compellent for ESX env and it's not bad. Using VNX5300 for Hyper-V and I don't like the interface at all. Would be nice if everyone used the close to the same terms and process for connect hosts to LUNS.
- I think EMC has done a great job with the VNXe. This class of storage has exactly these things, SMB needs.
- Start using lower end SSD as either a replacement for FC or a tier between FC and higher end SSD. Better integration with array tiering and SDRS
- Provide some kind of guidance of how to set iSCSI securely up - no deep knowledge of protocols required but still demonstrating the way towards secure iSCSI provisioning.
- Provide discounts for PowerPath/VE given the exclusivity of EMC products used + cost of the arrays themselves.
- Sometimes the integration of "new" VMware Features like VAAI or not good implement.It doesn't make it easy to test and use them.
- Auto Tiering in HP EVA, VAAI Thin Provisioning in P4500, VAAI in HP EVA, SSD Offering in HP P4500 G2, Selecting the disk from shelf/bay and present it to a LUN and make a datastore (using cool GUI), HP P4500 comes with iSCSI, FCoE and NFS.
- Better visability into the whole stack of storage, servers, hypervisors, OS and applications. Right now seems that a lot of tools are good at one or two of these but not all of them at the same time
- PUT THE DAMN PRICE ON THE WEBSITE. Holy shit, I can price out a $10k server from Dell, or HP, or wherever, but want storage? Nooooooo, gotta call someone.
- Move things like FAST and FAST Cache to on board controllers within the systems. Maybe even virtual or physical appliances. Allow for in-line de-duplication, io optimization, and caching before the writes get committed to the array. Bring these things server side. Allow me to cache the reads and the writes within a containerized pod (a server, a rack, whatever) so that when the writes are actually committed, the only thing sent down the pipe is the the data which isn't already deduped. Do this for under $50 a year per virtualized system and that's one hell of a deal.
- Expand the monitoring of NetApp OnCommand Insight Balance. The ability to monitor network throughput through our switches so we can see latency that may be network related.
- NetApp's total "unification of everything" in 1 box, but at a cheap price for small business. EMC's VNXe unifies, and is cheap, but architecturally is multiple systems. I want the holy grail for cheap storage.
- Enable to use multiple RAID types inside one pool. Example mix RAID5 Flash tier with RAID10 SAS tier.

Comments