Unsolved
This post is more than 5 years old
48 Posts
0
1503
December 22nd, 2016 07:00
Can disk failure cause server to reboot?
Hello,
Earlier today, we had a 'predictive failure' from a PowerEdge 720xd server on one of the disks. About 1-2 hours later server rebooted and came back up in about 5 minutes.
I signed on to the server, but it was extremly slow and database team reported that they were not able to start SQL services running on the server. At this time I have noticed that disk now was in failed a failed state
This went on for about 45 minutes and suddenly server started functioning as normal again.
There is no memory dump, but can see the event id 41 in system logs.
In Dell OME, I have also noticed ID 2048 (failed disk) and after that I see some warning messages, 2346, 2057 and 2123.
I find it be coincidental for server to crash/reboot right after predictivve failure.
Any ideas?
0 events found
Latif_ee7dc4
48 Posts
0
December 22nd, 2016 07:00
Forgot to add, As soon as the server came back up after it crashed, hotspare rebuild has kicked in. To me this looks like the time it took to rebuild the hostspare was the time we were not able to do anything on the server.
theflash1932
11 Legend
•
16.3K Posts
0
December 22nd, 2016 09:00
Not normal for it to be crippled while rebuilding, but certainly degraded, which can cause some programs not to work very well - or at all. Settings in the controller can be changed to use MORE of the system resources to rebuild - default is 30% - if it has been set to higher than that, it could certainly explain it. Should really never be set to higher than 30%.
Latif_ee7dc4
48 Posts
0
December 22nd, 2016 09:00
It was set to default (30%) as this was one of the first things I have checked. I have also seen the following on OME Alerts logs;
Device failed: Physical Disk 0:1:3 Controller 0, Connector 0
Error occurred: Error on PD 03(e0x20/s3) (Error 02).: Physical Disk 0:1:3 Controller 0, Connector 0
Virtual disk degraded: Virtual Disk 1 (SQL Data) Controller 0 (PERC H710P Mini)
Redundancy lost: Virtual Disk 1 (SQL Data) Controller 0 (PERC H710P Mini)
Are these normal after/during drive builds?
theflash1932
11 Legend
•
16.3K Posts
0
December 22nd, 2016 10:00
At what point? Before the rebuild? During? If during or after, possible the disk is bad or you used a non-certified drive?
Latif_ee7dc4
48 Posts
0
December 23rd, 2016 00:00
Before the server rebooted by itself, disk was in 'predictive failure' state.
Everything happened after it rebooted itself.
All the disks are Dell certified disks.