Website on Plesk server suddenly started to show 500 error: AH10292: Invalid proxy UDS filename

Follow

Comments

319 comments

  • Avatar
    Martin Vogt

    Same here: I had to restore the entire server this morning. First action: I've disabled the auto-update under "Tools and Settings > "Update Settings" > and uncheck all "Auto install [...]". From now i will run the updates manually.

    1
    Comment actions Permalink
  • Avatar
    Luis Manrique

    Thanks for share

    0
    Comment actions Permalink
  • Avatar
    Scott Demaret

    I think I will put apache updates on hold forever after this little gem ;-)

    3
    Comment actions Permalink
  • Avatar
    Fran

    Fixed! Thanks for the quick fix ;)

    0
    Comment actions Permalink
  • Avatar
    Reinhold Kainhofer

    Using the suggested version="2.4.41-4ubuntu3" always gave me an error about the version not being available. The solution in that case seems to first check which versions are available locally and then downgrade to that version:

    apt list -a apache2

    Auflistung... Fertig
    apache2/bionic-updates,bionic-security,now 2.4.29-1ubuntu4.17 amd64  [installiert]
    apache2/bionic 2.4.29-1ubuntu4 amd64

     

    This gives the information that version="2.4.29-1ubuntu4" is available and can be used:

    export version="2.4.29-1ubuntu4"; apt-get install apache2=$version apache2-utils=
    $version apache2-data=$version apache2-bin=$version

    0
    Comment actions Permalink
  • Avatar
    Alessandro Carini

    It works for us: Ubuntu 18.04.6 LTS and Ubuntu 20.04.3 LTS, but issues like this must be prevented in first place, It seems that Plesk does not have a staging environment to test upgraded packages from distro in first place.

    Such a light-minded attitude from a professional vendor?

     

     

    3
    Comment actions Permalink
  • Avatar
    Dave Owen

    Thanks so much for the workaround.

    0
    Comment actions Permalink
  • Avatar
    Mr Roberts

    So will someone let us know, when this is fixed and to "undo" the workaround we just did to get the sites back online again?

    0
    Comment actions Permalink
  • Avatar
    Luke Douglass (Edited )

    initial comments on how to fix ubuntu 20.04 worked for me via terminal... I too was thinking my server had been compromised, checking history, logs, htop etc for malicious activity so glad it was a simple fix.

    This is why I hate system wide updates by automatically... Oden plesk has been ok overall but even this is a reminder about scheduling updates vs assumed trust.

    Oden, feature request (not the place I know) add ipset and some geoip ip sets to firewall app please, we do this manually to prevent hacks iptables+ipset with cron to add and check rules added after plesk changes adjust firewall rules to iptables.  (block email ports except customers geo zone vs whole of internet etc)

    0
    Comment actions Permalink
  • Avatar
    Alexander Ostapenko

    After "apt-mark hold apache2" I guess some manual changes needed to unhold after plesk maintenance release?
    Will you notify here all updates about this issue?

    0
    Comment actions Permalink
  • Avatar
    Fabian Rohr

    This cannot be emphasised enough:

    but issues like this must be prevented in first place, It seems that Plesk does not have a staging environment to test upgraded packages from distro in first place.

    Such a light-minded attitude from a professional vendor?

    A breaking bug like this is serious and I hope to see Plesk coming up with a statement regarding this and how they will make sure to prevent this in the future.

    2
    Comment actions Permalink
  • Avatar
    Pedro Clemo

    Thanks for the really quick fix, so glad this was at the top of Google.  The two line solution for Ubunutu 20.04 worked perfectly and brought websites back online almost instantly.

    0
    Comment actions Permalink
  • Avatar
    Manuel

    FastCGI workaround worked for me.

    Ubuntu 18.04.6 in Vultr

    Plesk Obsidian v18.0.38_build1800210913.22
    1
    Comment actions Permalink
  • Avatar
    Pasqi

    Thanks... now is time for coffee.

    0
    Comment actions Permalink
  • Avatar
    Grow

    I just ran the command given by @Reinhold Kainhofer and looks like another server of mine still working with the new update. am I right to assume that? I'm just worried to key in the work around as I have very very (almost no knowledge) working with CLI

    0
    Comment actions Permalink
  • Avatar
    Jimit Shah (Edited )

    Thanks a lot !!

    I am running plesk on 1und1 server and I face this problem today ... by executing your commands , I really get my server back....

    0
    Comment actions Permalink
  • Avatar
    Sachin Awari

    Thank you. It worked and saved me.

    1
    Comment actions Permalink
  • Avatar
    Stefano Gattuso

    thankyou for fast solution, this morning i wake up with this error.... 

    0
    Comment actions Permalink
  • Avatar
    WURK

    2021-09-28 07:00:12 status installed apache2:amd64 2.4.29-1ubuntu4.17
    root@plesk:~# export version="2.4.29-1ubuntu4"; apt-get install apache2=$version apache2-utils=$version apache2-data=$version apache2-bin=$version
    Reading package lists... Done
    Building dependency tree
    Reading state information... Done
    E: Version '2.4.29-1ubuntu4' for 'apache2' was not found
    E: Version '2.4.29-1ubuntu4' for 'apache2-utils' was not found
    E: Version '2.4.29-1ubuntu4' for 'apache2-data' was not found
    E: Version '2.4.29-1ubuntu4' for 'apache2-bin' was not found

    0
    Comment actions Permalink
  • Avatar
    Marcel Kallen

    WURK you are probably using the 20.04 instructions for 18.04 server

    -1
    Comment actions Permalink
  • Avatar
    Mathias

    Ubuntu 18.04 on AWS gives this error both with or without sudo for the following command:

    export version="2.4.29-1ubuntu4"; apt-get install apache2=$version apache2-utils=$version apache2-data=$version apache2-bin=$version

    Returns:

    E: Could not open lock file /var/lib/dpkg/lock-frontend - open (13: Permission d enied)
    E: Unable to acquire the dpkg frontend lock (/var/lib/dpkg/lock-frontend), are y ou root?

     

    0
    Comment actions Permalink
  • Avatar
    Beyonds

    Thanks it's working

    0
    Comment actions Permalink
  • Avatar
    Krisztian Kovacs

    Hi Grow,

    what helped me was to (all in Plesk UI):

    1. change to "FPM Application served by nginx" instead of Apache, in "PHP settings" for the domain.

    2. in "Apache & nginx Settings" switched off "Proxy mode".

    Now my sites are online.

    I hope this helps!

    Anyway the provided "solution" above is not a solution, just a workaround, I would still expect a fix with the next update.

    5
    Comment actions Permalink
  • Avatar
    Fabian Rohr

    If the workaround is not working for you, please read the previous comments before replying here.

    1
    Comment actions Permalink
  • Avatar
    Grow

    @Travis Ecclestone

    thanks man! I'm worried to key in anything because I dont understand what the commands are doing and I'm worried I might break my server

    0
    Comment actions Permalink
  • Avatar
    tiziano (Edited )

    on our website doesnt work the fix. 

    The following packages will be DOWNGRADED:
    apache2 apache2-bin apache2-data apache2-utils
    0 upgraded, 0 newly installed, 4 downgraded, 0 to remove and 4 not upgraded.
    E: Packages were downgraded and -y was used without --allow-downgrades.
    root@localhost:~#
    root@localhost:~# apt-mark hold apache2
    apache2 was already set on hold.

    any suggest please? thanks

    0
    Comment actions Permalink
  • Avatar
    VG (Edited )

    Automated script to fix that problem in all Proxmox containers. Could be interesting for someone.

     

    #!/bin/bash

    UBUNTU18="apt-get install --allow-downgrades --allow-change-held-packages -y apache2=2.4.29-1ubuntu4 apache2-utils=2.4.29-1ubuntu4 apache2-data=2.4.29-1ubuntu4 apache2-bin=2.4.29-1ubuntu4; apt-mark hold apache2"

    UBUNTU20="apt-get install --allow-downgrades --allow-change-held-packages -y apache2=2.4.41-4ubuntu3.4 apache2-utils=2.4.41-4ubuntu3.4 apache2-data=2.4.41-4ubuntu3.4 apache2-bin=2.4.41-4ubuntu3.4; apt-mark hold apache2"

    #update to fit your needs
    for n in `pct list | grep 'plesk' | cut -d ' ' -f 1`
    do

    RELEASE=$(pct exec $n lsb_release -- -r | awk '{print $2}')

    if [ $RELEASE == "20.04" ]
    then
    echo "=== PLESK $n ====";
    echo "Ubuntu 20"
    echo $UBUNTU20 | pct exec "$n" bash
    elif [ $RELEASE == "18.04" ]
    then
    echo "=== PLESK $n ====";
    echo "Ubuntu 18"
    echo $UBUNTU18 | pct exec "$n" bash
    else
    echo "=== PLESK $n ====";
    echo $RELEASE
    fi

    done
    1
    Comment actions Permalink
  • Avatar
    VDX Engineering

    If, after the downgrade of Apache, you're still having trouble, it might be that your SSL settings contain +TLS1.3. The downgraded version does not support that.

    0
    Comment actions Permalink
  • Avatar
    Mark Desborough

    Like everyone else - Thank you for the quick response to this. 

    0
    Comment actions Permalink
  • Avatar
    WURK

    Marcel Kallen Unfortunately not. So no prevail. Managed to get the most important sites up running by running PHP throuhg NGINX and disable Proxy Mode. Have a running ticket, as I have waaaaay too many Plesk servers running to answer all calls alone.
    Questioning how this have happened though as all my auto updates are intentionally disabled.

    0
    Comment actions Permalink

Please sign in to leave a comment.

Have more questions? Submit a request