0

We run an asp.net/sql server data collection website with a hand full of clients dumping data in and running reports. We moved to a new server (specs below) and have had issues with it freezing and having to reboot it a dozen times over the pass six months. The hosting company has mentioned possible causes (listed below) but cant give a definite answer on what is going wrong. They have offered to reconfigure how ever I like. We have benefited from having a much faster system and really dont want to get rid of the ssd's unless they are the issue. Two possible setup changes that I've talked with them about are also listed below.

Any suggestions on what maybe causing the freezing issue as well as suggestion on a new setup would be great.

My main questions are: Do SSD generally have problems running the OS & SQL Server on the same RAID Array? and Are the new SSD's still unrefined enough to be running in a production environment?

Thanks

Current:
Xeon Quad Core E3-1270 3.40 Ghz
16 GB DDR3-1333 ECC SDRAM
First Hard Drive: 120GB Intel SSD
Second Hard Drive: 120GB Intel SSD
Third Hard Drive: 120GB Intel SSD
Fourth Hard Drive: 120GB Intel SSD
SAS 4 Port RAID Card
Windows 2012 Standard Edition - 64 Bit
MSSQL 2008 Web Edition

Possible Causes:
Running Sql Server & OS on same RAID Array
OS Software Issues
Using SSD's
CPU Underpowered
Not enough RAM


Option 1
2x Xeon Quad Core E5-2603 1.80 GHz
16 GB DDR3-1333 ECC SDRAM
1 x 240GB Intel SSD - OS
3 x 1 TB SATA HDD (7200 RPM) - SQL Server
SATA 4 Port RAID Card
Windows 2012 Standard Edition - 64 Bit

Option 2
Dell PowerEdge E3-1270v2 3.5GHz 4 Cores
16 GB DDR3-1600 UDIMM
4 x 128 GB Samsung 840 Pro SSD
Add-in H200 (SAS/SATA Controller), 4 Hard Drives - RAID 10
Windows 2012 Standard Edition - 64 Bit
7
  • 1
    This is really quite broad and you haven't given us anything to work with. I would suggest hiring a consultant to help you with this issue as it's very custom to your specific environment. We can help you if you have something very specific that you're wondering about but there isn't much we're going to be able to do without a lot more information.
    – squillman
    Nov 1, 2013 at 19:08
  • Sorry I know it is very broad question. My hope was to give as much info as possible. My main question is: Do SSD generally have problems running the OS & SQL Server on the same RAID Array? and Are the new SSD's still unrefined enough to be running in a production environment?
    – JonDog
    Nov 1, 2013 at 19:12
  • 1
    Nothing you've listed would be a culprit for the freezes. Comb through the event logs for errors around the times of the freezes. Test the RAM if you have the ability to withstand downtime. Do you have a sysadmin to look into this kind of thing?
    – Wesley
    Nov 1, 2013 at 21:52
  • We really can't troubleshoot this for you remotely, it's like calling your doctor and telling them your arm hurts: Absent any additional information they have no way to know if you banged it on a door frame or severed it with a chainsaw. Review the suggestions you've been given here and perform additional diagnostics on your end to narrow the cause down. When you know the cause you can start thinking about solutions...
    – voretaq7
    Nov 5, 2013 at 18:56
  • @voretaq7 check samsmith post below.
    – JonDog
    Nov 14, 2013 at 16:52

3 Answers 3

1

If a reboot is reqd, it seems unlikely that the problem is the overall hardware config. Full reboots indicate something fundamentally wrong probably with the hardware (though possibly with the SW)

A good next step:

  • monitor free RAM on the system
  • monitor free disk space

What you are trying answer is:

  • Is the server running out of resources, causing SQL or the OS to be so bogged down that a reboot makes sense or
  • Has the hardware itself hosed up (implying that your provider needs to fix the hardware)

For example, if you cannot even rdp to the system it implies that the hardware itself has a problem.

3
  • I regularly check available resources (ram, disk space, and cpu) and seem to generally use no more than 30% with exception to when we run reports. When we run reports we may periodically jump to 50% for short periods of time.
    – JonDog
    Nov 1, 2013 at 19:17
  • @JonDog post updated Nov 1, 2013 at 19:21
  • Thanks for the advise. After strong arming them into doing further diagnostics they found a bad memory stick. After replacing it we haven't had anymore issues.
    – JonDog
    Nov 14, 2013 at 16:50
1
  • SSDs are typically fine in server environments, provided they are spec'd properly and are appropriate for the workload/application.
  • Hosting providers kinda suck. They usually deploy low-end hardware, as it's a business with tight margins in many cases. They'll resort to offering you a "chassis swap" if they can't find the root-cause of an issue.
  • White-box or home-brew server hardware usually lacks the insight into the OS/driver/hardware interactions that branded equipment can have. HP and Dell, for instance, will give you some idea as to why an OS crashes, or will have a watchdog timer in place. Also, the out-of-band management interfaces (IPMI, DRAC, ILO) will provide useful information.

Of the options you've listed, I'd ask for the Dell option #2, configured with Dell OpenManage agents.

0

Check your Event Viewer. You'll usually see errors in there that should give you a starting point for troubleshooting. If you haven't already, you may want to run some of Microsoft's 'Best Practices' tools. They'll help you find configuration issues.

Not the answer you're looking for? Browse other questions tagged .