OMV Server running hot after upgrade to 4.x

  • I upgraded my server from 3.x to 4.x and since the upgrade it has been running hot. I haven't added any new service after the upgrade that wasn't running on 3.x. The only other change I see is the 3x increase in the number of processes running.


    Attached are the graphs.


    Processes:
    Memory:
    Processor:
    Temperature:

  • look at commands top and journalctl, what is wrong. I did today upgrade too, and this helped me. FE. i had to start and then stop monitoring, to stop some error "rrdcached plugin failed to connect to rrdcached", than I had to add to samba ntlm auth = yes to be able to connect with my media player and xp.

  • look at commands top and journalctl, what is wrong. I did today upgrade too, and this helped me. FE. i had to start and then stop monitoring, to stop some error "rrdcached plugin failed to connect to rrdcached", than I had to add to samba ntlm auth = yes to be able to connect with my media player and xp.

    Any tips on what I should be looking at? I don't have outputs from before the upgrade so I won't know what it causing it. So far I can see ~3% CPU usage by the VBox plugin, but I've always used it.


    The only errors I see in the syslogs are:

    Code
    Jul  1 02:24:04 Vault PlexMediaServer: Error parsing XML.
    Jul  1 02:24:04 Vault PlexMediaServer: Error parsing XML.
    Jul  1 04:12:54 Vault cron-apt: TypeError: 'NoneType' object is not callable
    Jul  1 04:12:54 Vault cron-apt: TypeError: 'NoneType' object is not callable

    Off-topic question: How are you tracking system temperature?

    I use SNMP plugin and send my info to LibreNMS running in a Virtual Machine.

    • Offizieller Beitrag

    I upgraded my server from 3.x to 4.x and since the upgrade it has been running hot. I haven't added any new service after the upgrade that wasn't running on 3.x. The only other change I see is the 3x increase in the number of processes running.

    While the increased number of processes is interesting:
    With all other factors being equal (cpu fan speed, ambient room temperature), it's CPU load that would have the greatest bearing on the CPU's temp. In your case, your CPU is not being stressed at all with either OMV3 or OMV4. There's no reason why your CPU temperature should have jumped like that, with roughly the same load.


    Since actual CPU load hasn't changed, from OMV3 to 4, and the jump in CPU temp is massive; I'd be suspicious of the temperature data in the chart or (it would have to be a wild coincidence) something has happened to the CPU fan.


    Another possibility to look at is fan speed in BIOS, if your BIOS supports it. You might be able to pin the fan to full speed, in BIOS, and see what the results are. Pinning your CPU fan to a higher speed in BIOS might fix the issue. (On the other hand, even a slow fan speed should work in your case - your CPU is not working hard.)


    On the command line, see if the output of the following agrees with the charts.
    sensors
    ____________________________________________________


    As an example, with a load bouncing between 0 to 3%, the output of sensors on my i3 box is as follows:


    acpitz-virtual-0
    Adapter: Virtual device
    temp1: +27.8°C (crit = +105.0°C)
    temp2: +29.8°C (crit = +105.0°C)


    coretemp-isa-0000
    Adapter: ISA adapter
    Physical id 0: +34.0°C (high = +80.0°C, crit = +100.0°C)
    Core 0: +31.0°C (high = +80.0°C, crit = +100.0°C)
    Core 1: +33.0°C (high = +80.0°C, crit = +100.0°C)


    ____________________________________________________


    Depending on the generation of CPU you have, Intel processors self regulate for temperature meaning they slow down if they get excessively hot. On the other hand, anything north of 80C is not good for long life.


    (BTW - for followup - I'm out of town until next weekend, starting tomorrow.)



    • Offizieller Beitrag

    While you may have more tasks running, it appears that nearly all are sleeping which would have little effect on CPU load.
    _________________________________________________________________________


    It seems that you have a hot box, literally. I think your "temp1:" is you mobo temperture sensor, running at 158F. The inside of your server is, decidedly, "warm".


    Also, while warm, 67c and 72C CPU (152 to 161F) core temps are way under the 76 to 82C (169 to 179F) span shown in the chart. Did you look at the chart at the same time you pulled the sensors output? Did they match?


    Lastly, as it seems, your i7 CPU can endure higher temps than my i3 can.
    ___________________________________________________


    Still, I'd look at BIOS to see if the CPU fan speed can be increased. I'd also look at ways to vent the box, with an additional case fan if necessary, to get your server's overall temp's down. (Check to make sure that your case's vented surfaces are not blocked - this would include cases with vented bottoms that are sitting on carpet. Put a piece of plywood under it, or something else that's hard so the case sits on it's legs AND has a gap for air flow.) Getting a better CPU fan is not a bad idea either.

    • Offizieller Beitrag

    I still can't get on board with the idea that the motherboard temp and CPU (all of them), jumped roughly 20 degrees C, after an upgrade. (Unless, of course, something is restricting air flow to the case.)


    Just for testing purposes, if it was me, I'd rebuild OMV3 on a usb thumbdrive for side-by-side comparison. (But that's the OCD in me.)

  • The temp1 temperature is off between the graph and the sensor output by almost 20c but other temperature readings are close.


    If I am going to create a fresh copy of OMV, I'll probably start with OMV 4.x to see how that compares. Thanks for the tips and help.

    • Offizieller Beitrag

    The temp1 temperature is off between the graph and the sensor output by almost 20c but other temperature readings are close.
    If I am going to create a fresh copy of OMV, I'll probably start with OMV 4.x to see how that compares. Thanks for the tips and help.

    As I understand the start of this thread, the problem developed after an upgrade from OMV3 to OMV4 (Was it a command line upgrade?)


    Just remember that you could do a clean OMV install using a CD as the source, and put the installation on a USB drive, leaving your current OMV4 installation alone and intact. (You might need to change the boot order in BIOS to boot from USB.)


    Depending on what you boot - you could do a side by side comparison.


    Another factor to consider, if you did a command line upgrade:
    Something could have gone wrong with the upgrade. It might be worth the time to build a USB thumbdrive (with OMV4) just to check out this possibility.

  • Thanks for all the replies and tips. I finally decided to move my server from a laptop to a Dell T30 (hoping for better performance, less headache and overall less time fiddling with my setup).

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!