Sun Storage 7000 2010.Q1.1.0 Release

With our apologies for the delay in announcing it here, a new minor update to the 2010.Q1 software has been posted. Please read the release notes for the bugs fixed (and new features added) in this minor update, and happy upgrading!


So I had a very, very disconcerting problem upgrading my 7410 cluster to this release. I need to because I believe I'm hitting CR 6934730. In any case, I followed the cluster upgrade instructions and at the point I failed over to the newly-upgraded head ("A"), everything went TU. The BUI/CLI on A because inaccessible (the "emergency shell" was all that came up). Head B was now rebooted, and because the "Appliance Kit" was non-responsive, there was no way to fail back to Head B. Ultimately, the only resolution was to "reset to factory" by adding "-c" to the "kernel" line in the GRUB menu. I did this for both, reconfigured the cluster and imported the pools. One 20TB pool tool quite a long time to import. In the meantime, my shares were by no means available. More like, I experienced an unscheduled outage of about 6 hours. Avoiding these is exactly why we bought a cluster configuration.

Now, this is the second time this has happened. It happened on a much older release, before we went into production with this unit. One interesting note is that several hours after the cluster was re-established, I got an error in the "clustron" PCI card from Head B. That had never happened before. Perhaps, just maybe this is intermittently failing and causing the problem at the failover step. It is scheduled to be replaced Friday morning. I surely hope this is the problem, as general issues with updating AK revisions would present problems. I'm crossing my fingers, but felling twice-bitten...

Posted by Charles Soto on May 26, 2010 at 07:34 AM PDT #


Yikes -- that's definitely no good. Do you have a support call open on this issue? If so, can you mail it to me? You have a bunch of symptoms here that don't necessarily point to a single problem, but I would like to look at the data to try to understand what happened here...

Posted by Bryan Cantrill on May 26, 2010 at 07:40 AM PDT #

Oh yeah, the clustron is being replaced in the morning. Do you just need the SR number? And thanks. I'd like to not repeat an "up until 3a.m." sort of thing any time soon (except for fun)!

Posted by Charles Soto on May 27, 2010 at 06:08 AM PDT #

Well I guess there's no harm posting it here: SR 72979404

I've sent a service bundle from each head. Should be documented in the SR.

Posted by Charles Soto on May 27, 2010 at 06:11 AM PDT #

I upgraded to 2010.Q1.1.0 last week on our two 7110's. We hit a very strange problem though that some of the files on our shares became inaccessible. These were exclusively Excel spreadsheets being accessed via CIFS. A member of the Administrators group on the 7110 could access the files no problem but other Active Directory users who appeared to have valid ACL's on the files could not access the files - Excel reported that the file was "read-only or encrypted" - resetting the ACL's from the WIndows Explorer GUI appeared to fix the problem but it was very strangely and a little disconcerting.

The files affected were seemingly random - in different locations on the filesystem, different users involved - all very odd.


Posted by Alan Wilson on May 27, 2010 at 09:23 AM PDT #

I've updated my 7410 cluster with minor nuisances...hardware updates took longer than expected and SIM0 had to be reseated...but now two of three controllers have hung with no explanation. My 7110 pulled out of it after 5 minutes, but the 7410 had to be rebooted. Cluster takeover was fine, but there's nothing in the logs and no alerts were thrown. I had updated the 7110 to the earlier Q1 release with no issue. Now I'm wishing I hadn't downloaded the "minor release". Oh, and remote replication from a Q3 7110 isn't working to the 7410 cluster either. It worked from Q3 to the earlier version of Q1, but not this latest release.

Posted by Kim Hendricks on May 28, 2010 at 06:33 AM PDT #

Just so everyone knows: I am following up via e-mail with Charles, Kim and Alan. So if you do leave a comment here about an open support issue, be sure to leave a valid e-mail address!

Posted by Bryan Cantrill on May 28, 2010 at 09:13 AM PDT #

Post a Comment:
Comments are closed for this entry.

Official blog of the Oracle Fishworks engineering team.


« April 2014