-
Simultaneous System Shut Downs With No Warning
We have five systems running FCP connected to a SAN. Our problem started about 3 weeks ago. All five systems simultaneously shut down with no warning. It was not a power issue, as nothing else in the building was affected. Lights, other computers and monitors, coffee maker, all retained power. Our battery backups never kicked in either.
Last week we had two similar occurences. Three of the five systems simultaneously shut down, two days in a row at apporximately the same time. It was the same three systems both days.
We had our techs pour through console and system logs for all five machines and the SAN and cannot find any reason for the shutdowns. It seems to be a very random issue. We cannot determine if it is a SAN issue or system issue. Has anyone experienced anything like this?
-
12 Replies
-
What brand of SAN are you using?
-
Spooky.
Go to your energy saver preferences and click on the schedule button as you can set auto wake ups and shut down there. Perhaps something got written it when it shouldn’t?
Jeremy
-
We are on Apple XSan with two RAIDs. Using XSan Admin v.1.4.
-
I’ve checked the settings on the Server and all the systems and the Engery Saver preferences all look good. All are set to never to put the computers to sleep.
-
Did you check the ‘Schedule …’ tab in Energy Saver prefernces?
-
That’s bizzare… be sure to Post in XSAN forum too – there’s a lot of uber knowledgable people.
Are you using all Apple approved gear? Like a SANbox for your fibre switch? Does it have battery backup? Did the MDC(s) go down? You said you’re on Admin 1.4… but are all the boxes using Xsan 1.4 sofware, especially the MDC?
…even when we lose the entire SAN network for whatever reason… the edit machines don’t go down. But we stayed on 1.3 because it proved stable for us.
-
hey Flake, I got this email from an equipment dealer that I buy equipment from and maybe it pertains to you:
“Issue or symptom
We’ve been receiving more reports of a possible conflict between Xsan and QuickTime 7.1.3 and/or iTunes 7. The problems appear to span Xsan 1.2 to 1.4. We urge you to avoid updating to the latest iTunes and Quicktime until more information is available.
When the problems occur, clients lose access to the SAN volume. Restarting the client does not help. Intermittently, the volume may reappear.
Solution
Revert to iTunes 6.x and Xsan 1.3.”
Jeremy
-
We are using a QLogic Sanbox 5200 and XSan MDC Connection Kit, I believe both are apple approved. All of our systems, including the MDC, are now running XSan 1.4, they were at 1.3 during the first crash so we updated, but all systems were at 1.4 during the second two crashes. We had been advised not to update to 1.4 because of the same stability issues you mentioned, but after the first crash we updated everything thinking that was part of the issue.
The MDC never crashed. It has stayed operational durning all three occurences. We have checked the logs on the MDC as well, and not even those logs give us a hint of what is going on. The connected systems do not lose connection with the SAN. The SAN is mounted when the systems go down, and it re-mounts with no problems when the systems are re-booted.
All 5 systems and the MDC are on battery back-up. We have had power hits before where the backups have kicked in and we were able to properly shut everything down. These crashes do not seem to be power related at all, none of the breakers are being tripped, and everything else in the building retains power.
Today is the one week mark from the second crash, fingers are crossed… Thanks for the thoughts on this, I’ll post in the XSan forum today.
-
Yes. The schedule tabs on both the MDC and the systems do not show any events scheduled at all. During all three crashes, the MDC has been completely fine. It has stayed on, and is not logging any warning events at all.
-
Hey Jeremy,
That is something that we have been monitoring closely as well. None of the systems have been updated to iTunes 7 or the new QT. The oddest thing about this is that our connection to the SAN on all affected systems is fine. The SAN is mounted when the computers crash and re-mounts with no problems on the re-boots.
We were at XSan 1.3 on the MDC and all systems during the first crash and then upgraded to 1.4 hoping that would be part of the solution. All systems and MDC were on 1.4 during the second two crashes, but because the first crash occured while we still at 1.3 we haven’t felt we needed to revert back.
Thanks,
Flake
Reply to this Discussion! Login or Sign Up