Installing Ceph | Unable to correct problems, you have held broken packages.

H

hacman

Renowned Member
Oct 11, 2013
91
8
73
Newcastle upon Tyne, UK
Hi all,

Having a bit of an odd one here.

I've just rebuilt a node in my cluster, installing from scratch using the latest install media.

When trying to install Ceph Reef, the following is given. It seems ceph-fuse and ceph-common have already been installed from the squid repo?

Code: 
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
ceph-common is already the newest version (19.2.0-pve2).
ceph-fuse is already the newest version (19.2.0-pve2).
gdisk is already the newest version (1.0.9-2.1).
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 ceph-base : Depends: ceph-common (= 18.2.4-pve3) but 19.2.0-pve2 is to be installed
 ceph-mgr : Depends: libsqlite3-mod-ceph (= 18.2.4-pve3) but it is not going to be installed
            Depends: librados2 (= 18.2.4-pve3) but 19.2.0-pve2 is to be installed
E: Unable to correct problems, you have held broken packages.
apt failed during ceph installation (25600)

Is there any easy way to resolve this that the team recommends? Has something changed in the install media that would cause these newer versions to be present at install?

Thanks!
 
Do you remember what you had to do to upgrade it to squid?

I updated the repositories to ceph-squid to try, followed by `apt update`, then an attempt to install from the UI and from terminal second (with `pveceph install --repository no-subscription`), but both failed due to the deps..
 
I'm having the same issue, but a lot worse? You be the judge:
I installed PVE 8.3-1 fresh on 4 nodes and saw that since the last time that this cluster had been installed, the latest version was Reef-18.2. I've been a loyal paying member of the Proxmox community for +-2 years and only recently am attempting to take all of my experience and knowledge and build a software development company.
After the fresh install and install Ceph-19.2 this go'round, I updated my network to work with NTP+PTP with multiple PPS (/dev/ptp) devices and Ceph immediately started to act up. Furthermore, it seems that corosync for the Proxmox cluster is then spreading the "corruption due to timestamp misalignment" to other nodes before setting back on the original problematic node. I'm now stuck with 19.2 installed and no way (except to reinstall AGAIN (!@#$)) in order to downgrade the ceph for the cluster.
Any help would be greatly appreciated for all.
Thanks!
 

Attachments

Hello. I made a bugreport for this. This error happens, if you were on squid-no-subscription before, updated the pve host BEFORE installing ceph. This causes the pve-base packages of ceph (librados, ceph-common) etc be upgraded to ceph 19.x (squid). If you then switch back to reef-no-subscription, because you decide to install reef instead of squid, the packages do not get downgraded and this causes this error.

So best thing is to only update the host if you are sure that you are on the correct planned pve-ceph version, that you will use. Possible workaround might be removing the packages by hand, and installing them again with the older repo enabled.
 
I have a similar issue where during the installation of Proxmox VE 8.3.1, i get an error of :

Bash: 
dpkg: error processing archive /tmp/pkg/ceph-common_17.2.7-pve3_amd64.deb (--unpack): cannot copy extracted data for ./usr/bin/ceph-conf to 'usr/bin/ceph-conf.dpkg-new': unexpected end of file or stream errors were encountered while processing: /tmp/pkg/ceph-common_17.2.7-pve3_amd64.deb installation of package ceph-common_17.2.7-pve3_amd64.deb failed

Anyone has any idea on how to fix this through the installer? Since because of this,my Proxmox VE forcibly stops the installation and i dont know how to continue or retry it... I tried to manually install a ceph-common and it did install something but i guess since it gave me this error that the console on ctrl + alt + f3 does not sync stuff? I don't get it ...
 
You must log in or register to reply here.
Top Bottom