« Updated CLARiiON/VMware Applied Technology Guide and other key docs | Main | I probably shouldnt do this, but my proudest EMC moment »

September 16, 2009

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00e552e53bd288330120a5cb1072970c

Listed below are links to weblogs that reference A couple important (ALUA and SRM) notes:

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

PaulO

To clarify though (since i almost crapped myself at first read thru...)

From the Clariion and VMware integration paper:
"On a CX3 or earlier CLARiiON storage ststems, the Most Recently Used (MRU) or Round Robin policy must be used with failovermode=1"

So RR is ok in non ALUA modes on CX3 or under.

Too bad customers (like us) with somewhat newish CX3-XX's get screwed again (FLARE29 anyone....)

Krzysztof Cieplucha

I assume we will never see FLARE 28.5 or higher on any CX3 box, but is there any chance that future releases of FLARE 26 for CX3 will support (and when) SCSI3 reservation with ALUA?

DougJ

I second that Paul, wish those of us with CX3's were able to get newer Flare. Chad, any chance of that? Frustrating to have a couple recent CX3-80's and no longer get the cool stuff without going with a CX4.

Chad Sakac

Keep it coming guys - I'm pointing to feedback here to the CLARiiON org. Some stuff I understand why it would be very, very hard (ergo VM-Aware Navi and search), but I'm pushing hard for a FLARE 26 patch that adds this ALUA SCSI-3 reservation.

Believe it or not - I understand the frustration personally - I purchased 167 NS-20's (which have CX3-10s included) in 2009 to distribute to the field and partners :-)

Thank you for being customers, and please, keep the feedback coming!

Will

My question comes from a different direction, PowerPath/VE. Based on what I've read, PowerPath/VE was built on the basis of ESXi. This means that all management is remote from the physical. So, we need to maintain a separate management server, either Windows 2003 or RHEL, in order to manage our PowerPath/VE installs. Further, based on the implementation of PowerPath/VE, we need to maintain the root passwords for our ESX hosts on this management workstation inside an encrypted "lockbox." Unfortunately, this means that: 1. Someone getting into my management workstation is able to affect PowerPath on all of the ESX/ESXi servers managed by it. 2. The loss of my management workstation means a disruption in management of PowerPath/VE until I rebuild. ..... All of that said, do we really need to go forward with this architecture? Can't we put the management utilities on the physical ESX servers? Why are we purely building around the ESXi model? My VMware admin is asking if the additional licensing cost for vSphere to support PowerPath/VE is worth it, and another mangement level is not something he wants to hear.

Chad Sakac

Will - thanks for the question.

- PowerPath/VE works with ESX or ESXi

- no management is needed per se with PowerPath/VE. All path configuration is 100% automated, PERIOD. except the initial licensing, I would expect that many (most) customers would never use the rpowermt tool.
- but if you want to get some reporting, you use a "rpowermt" utility. This utility can be on any VM (and you then get all the VM HA goodness), it doesn't have to be a physical host at all. The reason for that management model is wanted to stay away from COS tools (for all the good reasons VMware is also going that way). Think of rpowermt as being like VMware's vMA (in fact, you could even install it IN the vMA). It interacts with PowerPath/VE via deep vmkernel API calls (ESX CIM and others).

- the "lockbox" is only there if you want to use it (again, this is the same model as VMware's vMA - you absolutely DO NOT need to cache the ESX passwords.

FYI,management of PP/VE in the near/mid term will be extended via a vcenter plugin in addition to the optional CLI tools provided via rpowermt.

Hope that helps!

Will

Chad,
Thanks for the response. In regards to using powermt/rpowermt, I use powermt a lot on other OSs during LUN presentation, troubleshooting, and documentation. Also, I'm a CLI and scripting guy in many ways. The extra layer (indirect/remote CLI) still concerns me somewhat, but I've gotten used to Navisphere CLI.

Do you happen to know what encryption level PowerPath/VE wraps around the communication between the PowerPath management server and the ESX host?

Thanks again.

Jay weinshenker

Chad, I'm not first and foremost a storage guy so forgive me for what might be a stupid question... I read over the post a few times. So if I'm on esx 3.5 with a cx-3xx it sounds like we stay with MRU and may end up with all our LUNs going thru the same storage processor after it comes back online (standby).

We have two ways to keep that from happening:

a. Upgrade to a cx4-xx and FLARE 29 at which point we'll be able to utilize RR (but only if on vsphere due ALUA) which will send xxxx (1000 by default) IOs down one path and then the other path giving us better performance than MRU but still not powerpath aka MPP performance or

b. Upgrade to Vsphere which will allow us (with ent plus licensing) to use powerpath/ve which can talk ALUA reservations in SCSI-2 giving us better performance (in this regard anyhow) than going from a cx-3 to cx-4+ RR anyhow

I ask because this statement really muddies that:

FLARE 28.5 and FLARE 29 are supported on CLARiiON CX4, but are currently not supported on the older CLARiiON CX3.  Ergo, a CX4 can support ALUA with vSphere, a CX3 cannot.

A cx3 can support ALUA (SCSI-2) under vsphere but you've got to use powerpath to do it.

I get that right?

We've budgeted to go to a cx4 with vsphere and powerpath next year but I want to make sure I understand. 
summary
cx3 / esx 3.5 NMP MRU
cx4 / esx 3.5 NMP MRU
cx4 / vsphere NMP RR
cx* / vsphere+powerpath true mpp

so in this context going to a cx4 w/o vsphere doesn't buy me anything. correct?

Ben

Are there any plans to support ALUA on the AX4?

Trevor

In short ... if we are using 3.5, then need to use MRU, or path thrashing can be an issue. In vSphere thorugh round robin or fixed can achieve path balancing that can survive a host reboot (even withouth powerpath installed).

is this correct?

seroquel medication

Very inportant notes I recommended this to all my friends.

Generic Viagra

I would like to appreciate the great work done You

Gopi

Recently, i have installed powerpath in esxi.Afterthat, i can sees all the paths are active in esxi.But, while running rpowermt host=host ip check_registartion command,it ask the lockdown password and esx username and password.After 20 minutes,it througout the powerpath not found error.

please help me.

Gevorg

Thanks for the post. But I am confused about ALUA and Reservation relations.Can you please explain why ALUA implementation depends on ALUA.Or give me some info about that issue?

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Working...
Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.

Working...

Post a comment

Comments are moderated, and will not appear until the author has approved them.

  • BlogWithIntegrity.com

Disclaimer

  • The opinions expressed here are my personal opinions. Content published here is not read or approved in advance by EMC and does not necessarily reflect the views and opinions of EMC. This is my blog, it is not an EMC blog.

Enter your email address:

Delivered by FeedBurner