StoreEver Tape Storage
1748198 Members
2717 Online
108759 Solutions
New Discussion юеВ

Re: MSL5026 + DP5.1 drive going offline

 
Remon van Gijn
Frequent Advisor

Re: MSL5026 + DP5.1 drive going offline

Thats one I do not understand.
I can add a device (the changer) but when selecting the device the only choise in the selection box is the changer:?:?:? one, there is no scsi available.

Please mind that the backups and chenger are working fine. What is a good coincedence is that the drive where the error reports are about, are on the same scsi bus as the changer.
The other drive where these errors are not reported about has the bus to it self.
D Wong
Regular Advisor

Re: MSL5026 + DP5.1 drive going offline

Hi Remon:

I talked to the HP people that work on the Insite Manger agents and they know of this problem as a bug in version 7.0.0.0. There will be a fix in the next full release version 7.1.0.0. I don't know exactly when this will be released but sign-up to be notified by email for this product at the support website.

The only monitoring agent that you need to disable is the SCSI monitoring agent. You should be able to monitoring everything else.

Dave W.
Jack Ramos
New Member

Re: MSL5026 + DP5.1 drive going offline

You can add the below parameter to your Fiber Channel agent registry key and it will disable agent polling to tape only, but allowing the agent to continue to poll primary storage. Please use Microsoft├в s precautionary steps when modifying your registry.

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\CqMgStor\CPQFCA]
"DisableFlags"=dword:00000001

I'd recommend that you disable tape polling on each SAN/EBS host, except for one host. Then set that one host agent to poll every 30 minutes (as opposed to 2 minutes). This should cut down the window of failure drastically, but still allow you to view statistical information on your tape drives from one host.

Thanks,
Ja
Remon van Gijn
Frequent Advisor

Re: MSL5026 + DP5.1 drive going offline

thanks Jack, Dave

Jack, I have a scsi connected MSL5026 not fiber. Does this work then as well (with which reg key?)

Seems we will mostly wait out the new agents indeed. I'll wait out for jacks input to above question and start giving some points , Yes ?
David Ruska
Honored Contributor

Re: MSL5026 + DP5.1 drive going offline

Remon,

The original problem with the 6.40 agents was found in a SAN environment, where a short timeout caused aborts which could cause robot hangs.

The current problem (with the 7.0.0.0 agents) has so far been found to also cause some robot issues. It has been found to happen in a SCSI library environment. We are not aware of any drive hang (or offline) problems.

Just to rule out any problems with the management agents, I'll let Jack comment on the best way to disable it.
The journey IS the reward.
Jack Ramos
New Member

Re: MSL5026 + DP5.1 drive going offline

I don't know if tape polling can be disabled in a direct connect environment. I'll check with our engineers and get back to you ASAP.

Thanks,
Jack
David Ruska
Honored Contributor

Re: MSL5026 + DP5.1 drive going offline

Let me update this statement:

---
The current problem (with the 7.0.0.0 agents) has so far been found to also cause some robot issues. It has been found in to happen in a SCSI library environment. We are not aware of any drive hang (or offline) problems.
---

More accurately:
---
There has been one problem found with the 7.0.0.0 agents that causes some robot issues, in a Fibre/SAN library environment. We are not aware of any drive hang (or offline) problems.
---

So, we're not aware of any SCSI connected library problems with the agents at the moment. However, that doesn't mean we can't disable them to see if they have any effect.

The journey IS the reward.
Jack Ramos
New Member

Re: MSL5026 + DP5.1 drive going offline

I understand that we currently have no way of telling the agent to distinguish between disk and tape in a direct connect (SCSI to Library) environment.

Jack
Remon van Gijn
Frequent Advisor

Re: MSL5026 + DP5.1 drive going offline

Well to close this off for now, at least untill new agents are out:

I removed only the remote storage agents form the Ctrl panel HP management console.

This has stopped the trap sending indeed, but in insight manager server I still see it happen as that polls the hw status more frequent then NNM. So it slips by NNM for now which is fine.

I will see if I can find any polling reg settings as jack discribed, but the biggest 'pain' is gone for now.

thks for all the help. points dealed, hope I've done it fair for my first time.