MSA 1000 Background Parity Initialization Question

[JonU]

Hello,

I am running ACU 7.20.19.0 in a 2-node Windows 2003 clustered environment (both nodes are running Windows Server 2003 Enterprise Edition SP1) and the storage system we are using for the shared disks is an HP MSA 1000 Firmware ver.: 4.32 .

Recently, I created a new logical drive from the pool of unallocated space in the array using the ACU. I did not encounter any problems at the time, however, when I first created the logical drive, I received a blinking status message # 785 "Bacground Parity Initialization is currently queued or in progress...If background parity initialization is queued, it will start when I/O is performed on the drive..." I figured this was completely normal.

Subsequently, I have completed a clustered installation of SQL Server on the disk and I am not experiencing any problems.

My question is this: Why would I still see the status message # 785 nearly a week after the drive has been in use (reading and writing data)?

Should I be concerned that the parity information is not being written correctly and in the case of a failed disk I could lose my data? How long should this normally take? Does anybody know if there is a way to force the parity information building to complete?

Thanks in advance,

Jon

[Jefferson Humber]

Jon,

Interesting... I assume your seeing this message within ACU, does anything appear on the MSA LCD ?

I've found another post similar (same message), but not quite the same. Not definitive answer here either. -> http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=995234

[Claude Legare]

I run ACU 6.45.1.0 with msa1000 firmware 4.24 and I have the same scenario. All of the raid I configured, 6 months ago, still had this message last month. I didn't look at situation for a couple of weeks and now that I read your message, all alerts have gone.
It took about 3 to 4 months to complete but I think that it is because we added many units at the same time and I/O didn't happen immediately. I hope this will make you feel more confident on your msa..!

Claude

[JonU]

Thanks guys for the information. I guess I will wait for a while longer and see if it goes away.

Jon

[John Kufrovich]

Jon,

If you have the MSA cli cable you can issue a few commands to see what the actual status is.

>show units
or
>show unit #(LUN number)

If parity is still going on you will see it reported on each LUN.

>show this_controller

This will report which LUN the controller is currently working on.

[JonU]

Thanks John!

I am not quite sure what the CLI cable is. I know how to use the ACU CLI, is this what you are talking about?

Unfortuantely, our hardware is hosted offsite so it is not easy for me to get in front of the MSA physically. Is there a way to issue these commands remotely?

Thanks,

Jon

[Jefferson Humber]

Jon,

I'm not aware of a way to issue the CLI commands remotely. Some O/S's support the commands to a degree (i.e OpenVMS), but the complete set is not available still.

[JonU]

Ok Thanks Jefferson.

Jon

[Simon Grant IRL]

I have encountered the same error message on a MSA 1000 volume. The message "#785 Background Parity Initialization is currently queued or in progress" has been showing in the Array Configuration Utility for about 4 months now.

The use of the drive has increased lately and we are now expierencing poor performance on this drive. Other array's within the SAN are not affected.

Has there been any further information on this?

Thanks

Simon

[Scarletti]

I've just seen something on the HP site that says the parity information only gets written when the controller is inactive. In a clustered environment, it's never inactive because the clustering software is always polling the controller.

Have a look at this document:-
ftp://ftp.compaq.com/pub/products/storageworks/msa1000/MSA1000BPv42.pdf

[Jon Morlan]

Iâ m adding this update to this old thread since I believe that I have found the solution to this problem. I have been able to clear this error by doing the following, and since I could not find this process anywhere on the net; I figured I would add it here so that the next person googling would find this and be able to fix their issue.

The cause for this is that the Parity initialization is hung or frozen and some array tools do not show this since the Parity Initialization percentage complete is not shown.

I had several servers all affected by this issue and was able to clear them all be doing the following

Have Array configuration utility version 8.28.13.0 installed. This is cp010232.exe

From inside the array configuration utility, if you choose the logical drive that has the 785 alert on it and choose â More Informationâ you will see that there will be a percentage showing as to the percent complete. It seems that this process for me anyway was hung at 0 percent for several servers and 17 for another.

Other versions of the array config utility for some reason do not show this percentage readout, only that the status is â in progressâ . These versions include 8.50.5.0 and 8.50.6.0.

What cleared this issue here was a reboot. After a reboot on the affected servers was done the parity initialization took right off and ran to completion. I was able to watch this occur by refreshing the % readout from inside the 8.28.13.0 version of the ACU.

Bottom line here is that the Parity Initialization may be stuck for some unknown reason and since the only version of the ACU (8.28.13.0) that I could find shows the percentage complete. If you donâ t have a version of the ACU that shows this percentage readout it could be hung and you donâ t know it. And all it takes to resolve it is a reboot.


Searchable Keywords
Morjo15 morjo15 Array Configuration Utility stuck hung 785 cp010232.exe 8.50.5.0 8.50.6.0 8.28.13.0 solution fix