Problem with storage views

[ThibautL]

Hi,

Ineed your help because I can't find a way to have storage views working properly.

My config is :

• vCenter 4.0

• ESX 4.0 with PowerPath/VE

• Clariions CX4 and CX3

When i go to Storage Views tab i have these errors : "storage service is not initialized" or "an internal error was encoutered while initilizing storage service".

No error in vpxd.log but i have isolated some in sms.log (attached to this thread).

Maybe i forgot to do something but i can't find any documentation for storage views.

Thank you in advance for your help.

[nirvy]

Hello,

Are you by any chance using SQL server / oracle for the vCenter DB

I fixed this just now by following this kb..

Not sure though if it was from the oracle jar file or just from restarting the vctomcat service! (might want to try restarting the service before you go through the rest of the steps)

Edit: just to clarify, I am using MS SQL Server backend, NOT oracle, but it still worked!

[ThibautL]

First of all, thank you for your help.

I tried what you said but it's still not working.

I also had problems with Overview Performance Charts but i think you have to wait at least one day because it's not working until it gathered enough datas.

After one day, it happens to work fine (at least for me).

[nirvy]

Well the only other thing i can think of is, I believe the storage management service is a part of the tomcat web service, and by default the web service is not started for some odd reason (can check this by browsing to your ESX host and attempting to use the web interface, should get an Error 503.

To fix that you need to follow these steps:

To summarise:

Check webservice is running: service vmware-webAccess status

Check the webservice configuration: chkconfig –list vmware-webAccess

Set the webservice to start at run levels 3,4 and 5: chkconfig –level 345 vmware-webAccess on

Start the web service: service vmware-webAccess start

I have an almost identical setup to you, Clarrion CX 3 and 4, etc.

This and my last post were the only two things i did to a brand new build of vCenter and this just worked for me. I had the vCenter running for about an hour before I got both the Performance and the Storage Views working.

Even if this post doesn't help, it is still handy to have the web service working!

Hope this helps

[ThibautL]

Good to know for the webAccess. I will post again in few hours to tell you if it's working or not.

[ThibautL]

no luck on that one too. Still got the same error about initiliazing storage service.

[admin]

The log indicates that the storage monitoring service detected a duplicate scsi path id. Can you please do the following so that we can investigate further:

1. Open the file <vmware-installation-dir>\Infrastructure\tomcat\webapps\sms\WEB-INF\classes\log4j.properties.

2. Uncomment (by removing the # in front of each line) all the entries under the section "iBatis configuration".

3. Save the file and restart the "VMware VirtualCenter Management Webservices" service from the Windows Services console.

4. Upload sms.log and sms.log.1.

Thanks!

Pushkin

[ThibautL]

Ok , i attached the logfile.

As you said, It seems to be a problem with a SQL statement and a duplicate INDEX based on SCSI path that should be unique.

[admin]

Hi,

The SMS logs indicate that the query retrieving SCSI paths from VC might be misbehaving. Will it be possible to provide the following information:

1. How many SCSI paths do you have in your environment. The query reports 1043 paths. Does it sound right? A ballpark estimate is good enough.

2. Are you using PowerPath and NMP plugin to manage path to the array?

3. If you are using PowerPath, can you report whether you see an "*" (asterix) under the "Preferred" column in "Manage Paths" Dialog box for the datastore managed by PowerPath.

Thanks.

[ThibautL]

Thank you for looking at my problem.

1. VI Client shows 128 paths (on each HBA) , 35 devices and 17 targets.

2. I'm using PowerPath/VE for every datastore. Only thing that is not managed by PowerPath/VE is an ADIC Tape Library. (PowerPath have no licence as for now)

3. I can't see any preferred path, they all are active and "path selection" + "storage array type" are greyed outfor all PowerPath devices (not for NMP devices).

I should add that I have 2 PowerPath warnings on the ESX console about claiming path, this is a part of /var/log/vmkernel :

^{Jun 9 09:52:11 SCT-ESX4-010 vmkernel: 6:19:34:58.614 cpu6:4339)Cow: 2097: This is not a sparse disk : magicnumber 3499016243, version 4219207868}

^{Jun 9 09:52:11 SCT-ESX4-010 vmkernel: 6:19:34:58.623 cpu6:4339)VSCSI: 3411: handle 0(vscsi0:2):Using sync mode due to sparse disks}

^{Jun 9 09:52:11 SCT-ESX4-010 vmkernel: 6:19:34:58.623 cpu6:4339)VSCSI: 3469: handle 8258(vscsi0:2):Creating Virtual Device for world 4340 (FSS handle 131555)}

^{Jun 9 09:52:15 SCT-ESX4-010 vmkernel: 6:19:35:02.499 cpu0:4173)WARNING: ScsiClaimrule: 709: Path vmhba1:C0:T1:L0 is claimed by plugin NMP, but current claimrule number 250 indicates that it should be claimed by plugin PowerPath.}

Oh and last thing, this is not a regular dual fabric architecture. It's a dual FC switch, dual attachment, one fabric. But do not be afraid, this is not a production environment.

[admin]

Thanks for the information.

As mentioned earlier, we strongly suspect the query to get incorrect data for ESX hosts with paths claimed by PowerPath. We are trying to recreate the issue internally but have run into some setup issues. I will update this thread of any progress.

[admin]

Hi,

VMware support will like to get in touch with you. Can you email your contact number and/or phone number at "mjoshi at vmware dot com".

Thanks.

[froulund]

Hi

Is there any news on this one?

I have the samme problem, i think, but without PowerPath!

Regards/

Henrik Froulund Hansen

[admin]

Hi Henrik,

We are still investigating this issue. Can you please upload the logs according to the procedure from my earlier post? We would like to confirm that you are indeed running into the same problem.

Thanks!

[froulund]

Hi

Here is what i get from the sms log 🐵

Thanks for your help

Regards/

Henrik

[admin]

Hi,

After going over your logs, this is what I found. There is a volume that is shared by 4 hosts and one of the host seems to report different size, than the rest of the three, for that volume. SMS will be unable to pick the correct instance of the volume in such a case and will result in failure to initialize. Ideally, all the hosts sharing the volume should report the same size for the volume.

The naa name of the scsi volume is 0200080000600508b4001075280000700006ee0000485356323030. With the logs, I am not able to tell you which host is reporting the incorrect size. However, I can tell that for the same volume the sizes reported are: 536870912000 bytes (500 GB) and 644245094400 bytes (600 GB). Can you confirm that is indeed the case? Moreover, can you manually refresh storage on all the hosts and see whether that solves the problem. You will have to restart VMware WebServices to restart SMS.

Thanks,

Mayank

[froulund]

Wow you are fast! 🐵

Now it's running and all is green in my "vCenter Service Status" :o))

Thanks!!!

Regards/

Henrik Froulund Hansen

[ThibautL]

Did you see the same size mismatch in my logs ?

I tried to rescan & refresh the storage on each ESX then restart the vcenter webservice but nothing changed.

[admin]

Your problem is different. Powerpath plugin does not seem to report some fields that the SMS query depends on. I will probably have the setup issues with Powerpath resolved in the next day or two to work on your problem.

By the way, did anyone from VMware support contact you?

[Michelle_Laveri]

I have the same issue, since Monday when I installed PowerPath for VMware...

Done the rescan, and restart of vCenter services - but no affect yet....

Regards

Mike Laverick

RTFM Education

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

Author of the SRM Book: http://www.lulu.com/content/4343147