No "normal" update for Weeks now


(Dr Thomas Quinton) #1

For weeks now we do have a Problem with updating our Nethserver- looks like our database is broken.
Now we get following information in Cockpit:

failure: repodata/09c8eb4e98e39680b0bd7247048dd0a61460d057cd809990aaa473ed12a27984-filelists.sqlite.bz2 from extras: [Errno 256] No more mirrors to try. [Errno 12] Timeout on (28, ‘Connection timed out after 30000 milliseconds’) [Errno 14] HTTP Error 404 - Not Found [Errno 14] HTTP Error 404 - Not Found

any ideas?

(Dr Thomas Quinton) #2

in Nethserver GUI Information is:

Resolving RPM dependencies

(Dr Thomas Quinton) #3

Interestingly from console it is possible to do

yum upgrade

without failures

(Giacomo Sanchietti) #4

The log is relevant to centos mirrors.

I suggest you to do the following:

  • if a /etc/yum.repos.d/NethServer.repo.rpmnew exists, copy it over the real .repo:
    mv /etc/yum.repos.d/NethServer.repo.rpmnew /etc/yum.repos.d/NethServer.repo
  • try to clear the yum cache:
    yum clean all && yum update
  • if something goes wrong, check /etc/yum/vars/nsrelease is empty

(Dr Thomas Quinton) #5

obviously not existing!?!

mv: cannot stat /etc/yum.repos.d/NethServer.repo.rpmnew: No such file or directory

(Markus Neuberger) #6

Hi @iglqut,

I think this is an optional point, just proceed with cleaning yum cache…

All my Nethservers seem to be broken
(Dr Thomas Quinton) #7

OK- done- no updates at the moment, We´ll see.
Thank you so far.
I´ll respond after the next update!

(Dr Thomas Quinton) #8

OK update: from NS GUI update same problem but from cockpit it worked, very interesting- also interesting- cockpit tells you when to restart! Good feature.

The install/update may have failed due to metadata caching problem