1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Windows SOLVED Gremlins 3 - Windows Strikes Back...Again

Discussion in 'Tech Support' started by ModSquid, 8 Oct 2021.

  1. ModSquid

    ModSquid Multimodder

    Joined:
    16 Apr 2011
    Posts:
    2,360
    Likes Received:
    674
    Good morning one and all!

    Hoping someone can assist with another daemones in machina issue. It may be the proximity of Halloween, but one of the boxes is playing up and with increasing frequency.

    Symptoms are that it turns itself off for no apparent reason as if shut off at the wall, then reboots. It can sometimes then do this two or three times in a row before being stable for an unknown and varying amount of time. At first, I thought it was browser related, as it seemed to be happening when on the internet. However, the machine is mostly used for internet, so that's a poor evidence case, plus swapping browser didn't help and it also does it when just typing in Libre Office or snapping windows to the edge of the screen. It apparently did it once during a game as well, but this hasn't been as consistent an issue.

    I have been into the logs and found a LOT of yellow "distributed COM" warnings, but they don't coincide with the shutdowns and I've no idea what they mean. Text for these is like this (I've removed detail where coloured red as I've no idea if this is secure data):

    The application-specific permission settings do not grant Local Activation permission for the COM Server application with CLSID
    {25......54}
    and APPID
    {15C
    ......402}
    to the user
    xxxxx SID (S-1-5-.....-1001) from address LocalHost (Using LRPC) running in the application container Unavailable SID (Unavailable). This security permission can be modified using the Component Services administrative tool.

    What's also puzzling is that where I can find the usual "unexpected shutdown" errors, the timestamps don't align. The below example shows the log remarking that the previous unexpected shutdown was at 09:49, but the rest of the log clearly shows no issues (or entries at all, in fact) between 09:19 and 09:50:

    [​IMG]

    The full shutdown log text (with redactions) is like this:

    Log Name: System
    Source: EventLog
    Date: 06/10/2021 11:45:28
    Event ID: 6008
    Task Category: None
    Level: Error
    Keywords: Classic
    User: N/A
    Computer: ...
    Description:
    The previous system shutdown at 11:41:07 on ‎06/‎10/‎2021 was unexpected.
    Event Xml:
    <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
    <System>
    <Provider Name="EventLog" />
    <EventID Qualifiers="32768">6008</EventID>
    <Version>0</Version>
    <Level>2</Level>
    <Task>0</Task>
    <Opcode>0</Opcode>
    <Keywords>0x80000000000000</Keywords>
    <TimeCreated SystemTime="2021-10-06T10:45:28.1117700Z" />
    <EventRecordID>14006</EventRecordID>
    <Correlation />
    <Execution ProcessID="0" ThreadID="0" />
    <Channel>System</Channel>
    <Computer>
    ....</Computer>
    <Security />
    </System>
    <EventData>
    <Data>11:41:07</Data>
    <Data>‎06/‎10/‎2021</Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Data>9</Data>
    <Data>
    </Data>
    <Data>
    </Data>
    <Binary>E5070A00030006000B0029000700F800E5070A00030006000A0029000700F8003C0000003C000000000000000000000000000000000000000000000000000000</Binary>
    </EventData>
    </Event>

    I'm also seeing the weirdness of the timestamps on actual shutdowns not tying up - it might say "prev unexpected shutdown at 10:32" for example, but the ACTUAL shutdown record in the logs might be at 10:23, so I can't account for the slippage. The restart in this next example shows a supposed shutdown at 10:30 but the only error record around (but not at) that time is to do with memory, the text for which reads as "The system firmware has changed the processor's memory type range registers (MTRRs) across a sleep state transition (S4). This can result in reduced resume performance."

    [​IMG]

    I've ordered some new memory sticks in case it's that causing the fun, have reduced the base clock to 100 (it was auto-o/c'd before to about 103, but I know that can cause issues so have backed it out) and nothing else is really too far out of the ordinary that I can think of.

    Does anyone have any clue at all what might be causing this and a recommended fix?

    Thanks in advance, gang!
     
  2. Midlight

    Midlight Minimodder

    Joined:
    6 Jun 2011
    Posts:
    333
    Likes Received:
    175
    I think the first thing I would be checking would be the power supply.
     
    ModSquid and Flibblebot like this.
  3. ModSquid

    ModSquid Multimodder

    Joined:
    16 Apr 2011
    Posts:
    2,360
    Likes Received:
    674
    Oh! That's interesting - never even considered that. Is there a way to check PSU health, with a multimeter or similar, or just a case of swapping it and seeing if it's then stable?
     
  4. Midlight

    Midlight Minimodder

    Joined:
    6 Jun 2011
    Posts:
    333
    Likes Received:
    175
    The best way is swapping it out with a spare or from another system that you know works. Provided you have access to one.
    Using a multimeter is not really viable, the issue could be a momentary fluctuation on one of the voltage rails and unless you are measuring that one at the right time you would never see it.
     
  5. ModSquid

    ModSquid Multimodder

    Joined:
    16 Apr 2011
    Posts:
    2,360
    Likes Received:
    674
    Ah, okay. Cheers bud - I'll source a replacement and see if that works. Appreciate the input!
     
  6. ModSquid

    ModSquid Multimodder

    Joined:
    16 Apr 2011
    Posts:
    2,360
    Likes Received:
    674
    Think you might have solved this one, bud - many thanks. Swapped the PSU for a different one and it hasn't happened since (touch wood!). Case closed!
     
  7. adidan

    adidan Guesswork is still work

    Joined:
    25 Mar 2009
    Posts:
    19,371
    Likes Received:
    5,113
    Yeah funky power issues are usually the PSU.

    Obviously not always, sometimes you can just look at a pc the wrong way or it decides it doesn't like where you bought that sata cable from and spin a wobbly.
     
  8. Kernel

    Kernel Likes cheese

    Joined:
    29 Sep 2003
    Posts:
    1,193
    Likes Received:
    44
    These are just noise, they'll have no bearing on your issue. (Mostly replying for info for others)
    https://docs.microsoft.com/en-US/tr...gement/event-10016-logged-when-accessing-dcom

    Glad changing the PSU helped.
     

Share This Page