Question Upgrade Virtuozzo container from CentOS 7

[Dave W]

Actually, that doesnt seem to be the issue, just reading the code and the error reported;

Traceback (most recent call last):
  File "/root/almaconvert8-plesk", line 700, in <module>
    args.func()
  File "/root/almaconvert8-plesk", line 260, in process_cts
    results = pool.map(process_single_ct, args.CT)
  File "/usr/lib64/python3.6/multiprocessing/pool.py", line 266, in map
    return self._map_async(func, iterable, mapstar, chunksize).get()
  File "/usr/lib64/python3.6/multiprocessing/pool.py", line 644, in get
    raise self._value
  File "/usr/lib64/python3.6/multiprocessing/pool.py", line 119, in worker
    result = (True, func(*args, **kwds))
  File "/usr/lib64/python3.6/multiprocessing/pool.py", line 44, in mapstar
    return list(map(*args))
  File "/root/almaconvert8-plesk", line 547, in process_single_ct
    new_tmpl_list_raw = sorted(set(subprocess.check_output(['/sbin/vzpkg', 'list', ct]).decode('utf-8').split("\n")))
  File "/usr/lib64/python3.6/subprocess.py", line 356, in check_output
    **kwargs).stdout
  File "/usr/lib64/python3.6/subprocess.py", line 438, in run
    output=stdout, stderr=stderr)
subprocess.CalledProcessError: Command '['/sbin/vzpkg', 'list', 'xxxxxxxxx']' returned non-zero exit status 28.
Failure running almaconvert8 - Exiting... to resume use --convert and --finish options

/sbin/vzpkg list xxxxxxxxx

works ok from cli though, more digging required methinks

 

[websavers]

OpenVZ 7.0.21 - I suspect almaconvert8 only works smoothly on 7.0.20 and newer. That would explain some odd issues like this!

 

[Dave W]

Thanks, just updated the platform so now we're on VZ 7.5.6, latest release, but still getting the same error above.

More digging needed.

 

[websavers]

@Dave W Wish we could help more with this, but almaconvert8 is entirely written by Virtuozzo devs. The -plesk modified version does only one thing: excludes Plesk from the list of exceptions that would cause it to not run at all from the start.

Considering you appear to be running Virtuozzo direct, I'd recommend reaching out to Virtuozzo support and give them that stacktrace (but remove the -plesk from each instance of almaconvert8 as they're not going to want to help with that if it's been modified, I'm sure).

You might also try using another container to see if it gets different results, or even try it on a non-Plesk container using stock almaconvert8 to see what the results are.

Once that's fixed up, if you're able to identify issues in our wrapper (not in almaconvert8 itself), please create an issue on the GitHub project and we'll see what we can do!

 

[websavers]

@Dave W Just realized - did you remove /root/almaconvert8-plesk after updating Virtuozzo/OpenVZ? Our wrapper will likely need to copy an updated version of the software over and modify it, but it assumes you started this from an updated node.

 

[SteveITS]

try it on a non-Plesk container using stock almaconvert8 to see what the results are

FWIW it works on a slave name server container, I upgraded two. I just had to re-save the existing firewall config in Virtual Administrator afterwards, to get it to allow DNS. Have not tried a Plesk container yet.

(I also upgraded a non-VZ VM under Hyper-V by upgrading in two steps to CentOS 8 then Alama, and had to create firewalld rules there for DNS/SSH)

 

[Dave W]

Greetz, got it to work.

Some of the steps I took;

1. Make sure VZ is up to date
2. Refresh the VZ OS template caches
3. Upgrade MariaDB in the container to version 10.11, used Plesk for this, worked seamlessly, kudos to the devs
4.

yum remove nagios* nrpe

these caused dependency conflicts during the upgrade, easier to add afterwards.
5. I had to remove the -d option from snapshot creation in the almaconvert8_plesk script, it seems in Virtuozzo

--decription

not

-d

is allowed for the

vzctl snapshot

command. This meant that the script could not start/stop the container. The script seemed to pause for a timeout after removing Plesk components and dependencies so it seems that

        try:
            try_stop_ct(ct)
            snaphost_out = subprocess.check_output(['/bin/prlctl', 'snapshot', ct, '-n', 'Pre-Almalinux8', '-d', 'Created automatically by almaconvert8 tool at %s' % now])
            try_start_ct(ct)

failed through to

        except:
            snaphost_out = subprocess.check_output(['/sbin/vzctl', 'snapshot', ct, '--name', 'Pre-Almalinux8', '-d', 'Created automatically by almaconvert8 tool at %s' % now])

I will look into this further as I run the convert again a few times.
6. I had to reinstall the PHP verions and some of the Plesk Extensions, but that was ok.

I'll document in more detail any issues I find on restesting. Many thanks to websavers for this.

D.

 

[websavers]

@Dave W would welcome a pull request with these changes over on GitHub if only to ensure it works in more scenarios!

Technically we could modify the --d value to --description in alamconvert8_plesk via sed, though I'd also recommend submitting that to Virtuozzo devs. If you have a vz license, directly to them, otherwise to bugs.openvz.org as well, as that's their code!

The code is supposed to reinstall PHP 7.1 - 8.2 - hoping we can fix that too!

 

[Dave W]

Excellent, I'll try and get around to doing some more testing later in the week.