StoreOnce Backup Storage

HPE Storeonce 6200 - Slow Performance

 
SOLVED
Go to solution
Andy___H1
Occasional Advisor

HPE Storeonce 6200 - Slow Performance

Hi,

We have an old Storeonce 6200 backup device which has some issues, hoping you can advise. We have raised a case with HP for these issues, but thought we'd raise here also just in case anyone has any ideas.

We have source and target replication 6200 D2D's. For our target D2D, we noticed a few weeks ago it was getting behind on housekeeping and replication, and even local backups which happen in our target environment were also running very very slow. No errors in logs as to why so slow, D2D was running approx 90% full, and has been that full for well over 2 years now, running between 80-95% full in that time. But this was the first time all activities on this target D2D was slow. First fix was reboot, no change in performance.

So we moved the local backups off to another D2D to reduce the load, but replication & housekeeping still not completing, and both running slow. So took it down to just either just housekeeping, or just replication running on their own for a day, and still slow processing on each. Case raised with HP, and we're waiting for a code drop to put on a QR release to clear the D2D completely and start all replication processing again. We only have a months worth of data on this D2D, so the data has technically expired our target D2D anyway, hence the choice just to QR it. Given we can't send support ticket logs to HPE (customer restrictions) their hands are tied to get this fixed through normal routes. 

We did get a temporary hpsupport password and try out some underlying commands, but none suggested any issues, hence the choice to QR it. However, whilst waiting for the QR code to be delivered we thought we'd try something else, just delete all CIFS/NAS shares, i.e. all data, and see if we can just re-replicate and it will work again at normal speed. Issue is we've now deleted all data, so no user data at all on the D2D, rebooted, but the "Couplet Total Free Space" on the "Storage Report" on the GUI does not equal the "Couplet Total Capacity". i.e.:

Couplet Total Capacity = 121,413 GB

Couplet Total Free Space = 28,166 GB

Service Set Used Space = 0 GB

So we have no used space, but the free space doesn't equal capacity. I assume as we deleted the dozen or so CIFS/NAS shares it may be reclaiming the space in the background maybe? Does it take time for the Free Space to comeback? Or is this whole source of our performance issues, the fact the D2D can't seem to report and manage it's space usage causing it to run slow, and a QR update should clear it back to defaults?

Lots of questions, not many answers I'm afraid so far. Also we are getting new D2D's in soonish to replace these aging 6200's, but for now we don't have the new ones, we need to get this target one back up and running to normal speed.

Any help or advice would be gratefully received.

Thanks,

Andy

2 REPLIES 2
Amarjit_Y
HPE Pro

Re: HPE Storeonce 6200 - Slow Performance

Nature of issue is related to Performance.

required deep drive into the various support ticket and logs .

Please open a support case in order to proceed further on this issue.

Thank You!
I am an HPE Employee
Andy___H1
Occasional Advisor
Solution

Re: HPE Storeonce 6200 - Slow Performance

Hi,

As per my first statement, case has been raised. Was just hoping someone may have seen these issues before and could give some advice.

Anyway, as per my post, we wiped the D2D last Friday. Over the last weekend since I posted the space usage went back to normal, free space = total capacity, with no user data on there. So we've now initiated replication again from our source 6200's, and replication performance is back to normal speeds, and housekeeping is running as normal also. We have over 600TB of original source user data to replicate across, done about a thrid so far. Hoping the performance stays to acceptable levels once the D2D gets full again. This target D2D was never sized properly from day one years ago, and now it does get very full, which is not good for a D2D, but we don't have a way around this until we get our new D2D's onsite giving us more capacity.

So for now, we're back to normal, shame it took the actions of wiping the D2D of it's CIFS/NFS shares to get us there, but hey, we're just happy the D2D is usable again now as a remote target.

Thanks,

Andy