Test plans stops with "Array with key 'BB0050569F23260000-null' not found" error

[asmolenko]

Dear Gurus, I hope someone can help me with my issue.

My SRM 4.0 lab:

• two vCenter v.4 servers + SRM 4.0 + EMC_Celerra_Replicator_SRA

• two VSA Celerra with working replication

• two ESXi 4 servers connected to the vCenter servers

When I'm trying to execute a Test recovery plan, I've got the error "Array with key 'BB0050569F23260000-null' not found".

I assume that is someting with recovery VSA, but can't move further, since I am beginner in storaging.

I have attached screenshot from "Configure Storage Arrays" diag box, maybe it will help.

Thanks in advance!

[admin]

Hi Bill,

I am curious to hear what you will hear from EMC support. But it should be the same as I am saying. The updated SRA should fix the issue. And the update SRA should be out soon. Very soon hopefully!

Michael

[Bill_Oyler]

Michael,

EMC support has gone silent on us for the last week. Do you have any updates on your end? Do you have any contacts at EMC I could get in touch with regarding this issue? If so, could you send me a private message?

Thanks,

Bill

Bill Oyler

Systems Engineer

St. Croix Solutions

www.stcroixsolutions.com

[admin]

HI Bill,

I have checked on the status of this and quite frankly, there are some technical issues. We are working on it, and as soon as we can we will get the SRA out. There is nothing more we can do to make this go faster.

Michael

[Bill_Oyler]

Just wanted to give everyone an update on this issue. EMC support shared a new version of the Celerra SRA with me: v4.0.17. For whatever reason, the most current version available from vmware.com is 4.0.13 (Released 10/05/2009). (I'm not sure what happened to all the versions in between, and why they aren't available to download.) At any rate, using the 4.0.17 version of the SRA, I was able to completely solve the above problem with my Celerra Simulators in my lab. Now SRM 4.0 works perfectly in my lab!

However, with the physical Celerras I am working on in a different environment, I am now having a completely different problem. When adding the primary Celerra to SRM with the Array Manager Configuration in SRM, I get the following error:

Invalid SRA response. discoverLuns for array XXXX returned NFS volumes but discoverArrays did not return any NFS server addresses.

This error doesn't make any sense because we are not using any NFS datastores. There are no NFS exports on the NS480. This is a 100% iSCSI environment. So the SRA v4.0.17 seems to have solved one error message but introduced another one, at least in the case of this NS480 replicating with an NX4.

I still have a ticket open with EMC so I will report back any findings. It sure seems like the version 4.0 Celerra SRA is a lot more fussy than the version 1.0 SRA was.......!

Bill Oyler

Systems Engineer

St. Croix Solutions

www.stcroixsolutions.com

[admin]

Hi Bill,

Thanks for the update. I am glad you will keep working with EMC. I have heard you comments about this SRA elsewhere so with you sharing your issues with EMC we all will benefit with a better SRA!

I want everyone who reads this to understand the only SRA's people should use is the one they download from VMware.com. The reason for this is that the SRA's, once they are finished by the vendor, are sent to VMware for test and certification. If you don't see it on VMware.com that means we think it is not ready yet.

I still do not see inside of VMware the Celerra SRA but as soon as it is available, or I know a reliable date I will share it in this message thread.

Thanks Bill for working with EMC on this - I most appreciate this!

Michael

[Michelle_Laveri]

Like others here I got a new version of the SRA for lab environment. Unfortunately, my problems persist. I suspect the same error message has a source of more than one cause. I noticed with interested that Chad Sakac on his blog post about the new Celerra VSA might need an updated SRA...

"For those using the new Celerra VSA – updated docs and videos are going up, but it works easily – you just don’t need eth2 anymore. eth0 and eth1 are all that is needed. the “control station” access comes from those interfaces, no need for eth2. The new Celerra VSA is faster, works in less memory, and everything is working great… one exception – rarely SRM issues – we think this may need an SRA fix – working on it."

http://virtualgeek.typepad.com/virtual_geek/2010/01/breaking-a-cardinal-blog-sin.html#more

Regards

Mike Laverick

RTFM Education

http://www.rtfm-ed.co.uk

Author of the SRM Book:http://stores.lulu.com/rtfm

Free PDF or at-cost Hard Copy

[bladeraptor]

Hi

I am writing this as an EMC employee.

Everything had been working fine in my lab with two physical Celerras.

I made the upgrade to vSphere U1 and on advice conducted a 'repair' of my SRM installation.

Since then the iSCSI / NFS / Celerra SRA has been throwing up the error described in this thread.

I do not think it is the SRM installation that is the problem as both my RecoverPoint and CLARiiON deployments on the same SRM and VC environment work fine - although for both I did have to go into array configuration and run through the wizard to force SRM to rediscover the sotrage and then did a storage refresh

I have sent, as some of you on this thread have done, the Celerra SRA logs into support and I know the developers are working on it

What would be really helpful to understand and I will convey it back to the developers is whether any of you have ever had the Celerra SRA working fine with vSphere and if you have but it has now stopped working - do you know what changed?

For me the only change was the U1 update

Many thanks

Alex Tanner

[Bill_Oyler]

Alex:

Here is a recap of my various experiences with SRM and the Celerra SRA:

1) ESX 3.5, VC 2.5, SRM 1.0, Celerra SRA 1.0, with a pair of Celerra Simulators (original DART 5.6.36) = worked great!

2) ESX 4.0 GA, VC 4.0 GA, SRM 4.0 GA, Celerra SRA 4.0.13 (GA), with a pair of Celerra Simulators (original DART 5.6.36) = error message in subject line of this thread.

3) ESX 4.0 U1, VC 4.0 U1, SRM 4.0 GA, Celerra SRA 4.0.13 (GA), with a pair of physical Celerras (DART

5.6.42) = error message in subject line of this thread.

4) ESX 4.0 U1, VC 4.0 U1, SRM 4.0 GA, Celerra SRA 4.0.17

(unreleased new version), with a pair of Celerra Simulators (new DART

5.6.47) = works great!

5) ESX 4.0 U1, VC 4.0 U1, SRM 4.0 GA, Celerra SRA 4.0.17

(unreleased new version), with a pair of physical Celerras (DART

5.6.42) = new error message (*Invalid SRA response. discoverLuns for array XXXX returned NFS

volumes but discoverArrays did not return any NFS server addresses.).*

Hope this helps.

Bill Oyler

Systems Engineer

St. Croix Solutions

www.stcroixsolutions.com

[buckmaster]

Wondering if there is a fix? Having the same issue as the beginning of thread. Celerra vsa version 5.6.47-11, vSphere 4.01, Celerra SRA v 4.0 from VMware downloads, SRM 4.0. Trying to learn SRM, makes it tough when you can't run a test recovery.

Originally, I started out with the Lefthand Appliance and followed Mike L. instructions. Could run a "test recovery" plan however got a license error on remote copy when trying to "RUN Recovery" on the LeftHand VSA.

[admin]

Hello Buckmaster,

LHN will not support a failover with the no license operation. Test yes, failover no. The updated Celerra SRA is almost upon us. Real soon now.

Michael

[Bill_Oyler]

Just to keep everyone updated - EMC provided me with a new 4.0.18 version of the Celerra SRA, which enabled me to get further than ever before on the DR Test, but now the SRA hangs while the Celerra is creating a writable snapshot of the iSCSI LUN, causing SRM to time out. The EMC engineer working on the case has seen this problem before with another customer and suspects that it is another issue with the SRA. So I'm still holding out for the next SRA... I'll keep everyone posted when it finally is fixed.

Bill Oyler

Systems Engineer

St. Croix Solutions

www.stcroixsolutions.com

[admin]

Hello all,

I am happy to say that there is a certified version of the Celerra SRA that will avoid this error. It is 4.0.17 and, unfortunately, I am in the middle of rebuilding my EMC part of the lab, so I cannot say I have tested this SRA. So please let me kno whow it goes. This SRA is available via our SRM downloads page.

Michael

[Bill_Oyler]

Michael,

Is this 4.0.17 the same 4.0.17 that I tried a few weeks ago and got the above NFS error message? Or is this a "new" 4.0.17? The 4.0.17 that I used the other week mistakenly identified some CIFS file systems as NFS exports, and crashed SRM with an error message when configuring the Array Managers. The 4.0.18 version that I obtained on Friday fixed this issue, but it has different issues creating writable iSCSI LUN snaps. Is it worth my while to "downgrade" from 4.0.18 to 4.0.17, or wait for the next 4.0.18 fix?

Bill Oyler

Systems Engineer

St. Croix Solutions

www.stcroixsolutions.com

[admin]

Hi Bill,

I am not sure. I can say this is the first SRA that has passed since we started on this journey. So I would assume that it is a different one and things will work. However, I will wait to hear how it goes with you! Please let me know.

I will be irriated, as I am sure you will be, if it is the same one. But please confirm, and I will escalate.

Michael

[buckmaster]

It would be great if the file itself had a revision number built into the name.

[admin]

You mean other than the download file? I think that is pretty good. You should see some of the other names. I would like to see a readme file included. EMC does very good readmes, but they don't always end up in the SRA download or install.

[Michelle_Laveri]

Well, I'm trying the new SRA and getting the same error message as before.

I'm not really sure what the source of the problem is. Whether it's the internal name of the Celerra or the hostnames themselves or other objects. In fact I'm not really sure if is a character issue at all...

Regards

Mike Laverick

RTFM Education

http://www.rtfm-ed.co.uk

Author of the SRM Book:http://stores.lulu.com/rtfm

Free PDF or at-cost Hard Copy

[Michelle_Laveri]

I'm pleased to say I recently got a new SRA (4.0.18) and this error has gone away...

Regards

Mike

[admin]

Hi Mike,

Thanks for sharing this. It is not a certified or support copy of the SRA however. I have shared this back up the organization to see what is going on. I confirmed the latest certfied SRA is 4.0.17. Hopefully we can get the 4.0.18 version out soon, but of course we need to recieve it first!

Michael

[buckmaster]

Does anyone know when the new SRA is going to be released? 4.017 is still posted on the download site. Does anyone have the Celerra VSA working? I'm just trying to learn SRM not really the Celerra. Are there other alternatives to the Celerra VSA. I've tried the Lefthand unit and it works with test plans but it is cripled when you run a failover with a license validation. I know about Falconstor but the license is only good for 30 days. I really hate to install a temporary solution.

Thank you.