Hardware Thread, HP DL180 server in Technical; Hi
We've bought two new HP DL180 rackmount (2U) servers. One's been fine; the other seemed to randomly reboot itself. ...
12th September 2008, 01:08 PM #1
HP DL180 server
We've bought two new HP DL180 rackmount (2U) servers. One's been fine; the other seemed to randomly reboot itself. We had this one swapped out as I thought it was a hardware issue - there was no record of anything in the event viewer. However, now the new one's here, it's just rebooted itself without any warning! Again, nothing in the event logs. We had users logged in and using it (although as the reboot didn't take much time, I don't think they noticed - no phone calls yet!).
Anyone experienced anything like this?
IDG Tech News
12th September 2008, 01:11 PM #2
Have you checked/replaced the power lead and source? Are the servers both connected to a UPS?
12th September 2008, 01:13 PM #3
Yes, both the servers are on a UPS (and the previous one which we sent back was, too). I hadn't thought of the power lead itself, but will try replacing it at a convenient point.
12th September 2008, 05:49 PM #4
Do you have any USB devices in the server?
12th September 2008, 05:52 PM #5
the other thing is what are its settings for autoupdates?
12th September 2008, 07:07 PM #6
Are they on top of each other and one is over heating?
12th September 2008, 07:17 PM #7
Is your rack properly grounded. We once had issues with a faulty VDS system that was happily earthing itself into the wall cabinet, unfortunately the cabinet was not properly earthed and it was leaking back into the other components.
We got the dirty little VDS unit replaced and grounded the cabinet which solved the problem.
15th September 2008, 09:02 AM #8
Thanks for all the responses everyone; firstly, the rack is fully grounded so I don't think there's much of a problem there (nothing else in the rack - containing about a dozen servers - is affected). They are on top of each other but the room is fully air conditioned, so as far as I can see there shouldn't really be any overheating - but I will relocate the thermometer to the rear of the affected server for a week and see what readings we get.
The autoupdates are turned off (both by GP and manually, just to be sure...!), but I'm intrigued by the USB question. Currently there is no USB device on that server, but we do have a USB DLT backup drive which was attached to the former server (the one I had to send back); I've not yet connected the drive to this new server as I'm waiting for a new release of the software, which they assure me is due any day now (and has been for about a month!). As this drive has the option of connecting via USB or eSATA, I thought this time I'd try the eSATA route.
Any thoughts on any of this appreciated. Thanks again for all your comments and replies.
19th September 2008, 04:53 PM #9
OK, I think I've narrowed it down - it would appear to be a loading issue. This afternoon, for example, I was running a simple search for JPGs on the server (from a mapped drive, not actually on the console), and suddenly the server rebooted. Nothing in the event logs - you'd think it had never happened!
19th September 2008, 06:34 PM #10
Have you run a full scandisk of the drives, it is uncommon but perhaps there is some rather nasty unhanded corruption of the file system that drops the machine without warning.
Originally Posted by sdc
22nd September 2008, 01:40 PM #11
Full scan disk run; no errors reported. So, running a few tests, all seemed OK. Then ran a test backup on a selection of a few folders, using Windows NT Backup (Server 2003). Backup finished OK, but then the screen blanked (background remained, all icons and start menu vanished etc), the disk lights all turned on continuously. So I hit CTRL-ALT-DEL, before I could click Task Manager, the screen went black and it rebooted...!
Again, nothing in the event logs when it came back up.
23rd September 2008, 12:52 AM #12
Does it have the latest BIOS/Firmware installed for all of its devices, this can be a major cause of instability.
Also have you got any brother printers installed on it, their filthy drivers leak memory into the nonpaged pool and can cause crashes similar to the one you described. Check how much is in use by going into task manager, the processes tab, view menu, select columns and checking the non paged pool box.
Check through all of your processes to make sure that none are using up lots (more than 500kb) as there is only around 5mb allocated total. You would need to be looking just before a crash to detect it definitively unfortunately but if a process is using a lot it could be the problem.
24th February 2009, 12:03 PM #13
Just noticed that I'd not followed this up with the fix... so:
Turns out the server requires SP2 to be installed for Windows 2003. Having done this, all is fine!
(We now have several of these servers, all running fine, apart from one new one which appears to have a graphics card problem as it's blue screening and rebooting upon attempting to start the GUI after installing Windows apparently successfully...! But this can (and hopefully will!) go back...)
Thanks to everyone for your help/advice.
24th February 2009, 12:15 PM #14
So if it requires 2003 SP2, why didn't HP install it in the first place? And what specifically does SP2 contain that stops random reboots? Seems a little odd to me.
24th February 2009, 12:16 PM #15
And me... but we had an HP engineer on site for over a week trying to sort this! Eventually, after he made countless calls to various other HP guys, he came across this "fix".
Originally Posted by Michael
By cuke2u in forum MIS Systems
Last Post: 18th July 2008, 03:34 PM
By Zoom7000 in forum Windows
Last Post: 6th July 2007, 12:43 AM
By mrforgetful in forum Windows
Last Post: 17th June 2007, 02:51 PM
By ajbritton in forum Thin Client and Virtual Machines
Last Post: 31st August 2006, 07:19 AM
By pete in forum Wireless Networks
Last Post: 11th July 2006, 11:07 PM
Users Browsing this Thread
There are currently 1 users browsing this thread. (0 members and 1 guests)