[SOLVED] Can't edit anything under /etc/pve

[Matthew18]

Hi,

I am trying to edit some configuration files under /etc/pve but whatever action I try to run from root it always tells me "Operation denied".

I have not configured a cluster, that node is standalone and Corosync or any other cluster features was never used, what could cause this?

 

[dietmar]

what editor do you use?

 

[Matthew18]

what editor do you use?

I'm using nano to edit the files, also there was an issue with the server, which forced me to flash the BIOS and when I did that the date of the server was modified with a wrong one (around 2081) and now it shows that the files (like .clusterlogs) were modified in 1 Gen 1970, that's also preventing machine statuses from update, like for example, if I stop a machine, start it or create a new one, the icon won't update at all, it will just stay gray or won't even change the hostname in the list.

Is there any way to remove the readonly state from that folder? Because I am unable to do anything on those files

 

[Matthew18]

Look at these screenshots. (first one is the Cluster log and the second one is the VPS I have created right now, also that's not english, it would be "Jan 1" in the cluster logs)

that's referring to wrong dates.

Spoiler: Screenshots

 

[Matthew18]

As you can see also here, there are wrong dates to the files all over the place.

Spoiler: Screenshot

 

[dietmar]

Is there any data/VMs on this host?

 

[Matthew18]

Is there any data/VMs on this host?

yes, I have some clients already running on it, I can't reinstall it unfortunately, unless I maybe take a backup of everything, but isn't that possible to solve it without a reinstall?

 

[rhonda]

Just to be on the safe side: Can you check the output of "pvecm status" and send that along?

 

[Matthew18]

Just to be on the safe side: Can you check the output of "pvecm status" and send that along?

pvecm status
Corosync config '/etc/pve/corosync.conf' does not exist - is this node part of a cluster?
Cannot initialize CMAP service

that's the output

 

[Stoiko Ivanov]

Some files in /etc/pve having a date on 01.01.1970 is normal/expected behaviour (they have the same time on my system) - see https://pve.proxmox.com/pve-docs/chapter-pmxcfs.html .

which files do you try to edit? (some files are hardcoded to be readonly by pmxcfs)

does `touch /etc/pve/testfile` work? (does it create `/etc/pve/testfile`)

Setting the date to 2081 probably messed up the rrd-files for good (you won't get any updates in those file until after 2081) - the files get persisted in `/var/lib/rrdcached/db` - AFAIS - you could try to move them away (stop the rrdcached service, backup the whole /var/lib/rrdcached directory, move it away, try starting rrdcached again)

 

[Matthew18]

Some files in /etc/pve having a date on 01.01.1970 is normal/expected behaviour (they have the same time on my system) - see https://pve.proxmox.com/pve-docs/chapter-pmxcfs.html .

which files do you try to edit? (some files are hardcoded to be readonly by pmxcfs)

does `touch /etc/pve/testfile` work? (does it create `/etc/pve/testfile`)

Setting the date to 2081 probably messed up the rrd-files for good (you won't get any updates in those file until after 2081) - the files get persisted in `/var/lib/rrdcached/db` - AFAIS - you could try to move them away (stop the rrdcached service, backup the whole /var/lib/rrdcached directory, move it away, try starting rrdcached again)

RRD graphs are working fine after deleting those folders, but it's still not updating the status of the machines at all.

 

[Matthew18]

Hi,

I decided to reinstall the server, backing up all servers now.

 

[Matthew18]

Hi,

while making backups it gave me this error:

RPCEnvironment init request failed: Unable to load access control list: Connection refused (500)

-- EDIT --

now if I try clearing up the cache it says "ERROR 501" and won't load the page at all.

 

[Matthew18]

Hi,

a restart did the trick, everything is resolved now, thank you for the support.