in #41213 (closed), the gitolite server (cupani) was locked down by disabling the SSH command completely. if there are no complaints on April 24th, shutdown the server.
note that i've migrated the gitolite-admin repository to gitlab, as it's used to keep track of migrations and we want to keep updating it. it's not archived, but will be as part of #41215 (closed) once the migration is over.
that said, i've pushed an earlier version of the repo to cupani which was probably the last thing we needed to do here at all, before the retirement, although we might want to double-check the private repositories listed in #41215 (closed) are available on vineale before shutting down cupani.
messed up the retirement, because i didn't make the right schedules:
anarcat@angela:fabric-tasks[1]$ ./retire -H cupani.torproject.org retire-all --parent-host=fsn-node-01.torproject.orgstarting tasks at 2024-04-26 05:13:53.439705+00:00checking for ganeti master on host fsn-node-01.torproject.orgganeti node detected with master fsn-node-01.torproject.orgchecking on fsn-node-01.torproject.org if instance cupani.torproject.org is runningstopping instance cupani.torproject.org on fsn-node-01.torproject.orgWaiting for job 945564 for cupani.torproject.org ...scheduling cupani.torproject.org instance removal on host fsn-node-01.torproject.orgscheduling gnt-instance remove --force cupani.torproject.org to run on fsn-node-01.torproject.org in 7 dayswarning: commands will be executed using /bin/shjob 33 at Fri May 3 05:14:00 2024scheduling cupani.torproject.org backup disks removal on host bungei.torproject.org and director bacula-director-01.torproject.orgchecking for path "/srv/backups/bacula/cupani.torproject.org/" on bungei.torproject.orgscheduling rm -rf "/srv/backups/bacula/cupani.torproject.org/" to run on bungei.torproject.org in 30 dayswarning: commands will be executed using /bin/shjob 122 at Sun May 26 05:14:00 2024checking for path "/srv/backups/pg/cupani/" on bungei.torproject.orgpath /srv/backups/pg/cupani/ not found: [Errno 2] No such filescheduling echo delete client=cupani.torproject.org-fd yes | bconsole to run on bacula-director-01.torproject.org in 30 dayswarning: commands will be executed using /bin/shjob 65 at Sun May 26 05:14:00 2024Notice: Revoked certificate with serial 103Notice: Removing file Puppet::SSL::Certificate cupani.torproject.org at '/var/lib/puppet/ssl/ca/signed/cupani.torproject.org.pem'cupani.torproject.orgSubmitted 'deactivate node' for cupani.torproject.org with UUID f281de40-8419-43f2-9812-46021e173d64completed tasks, elasped: 0:00:33.910417 (user 6.71 system 0.19 chlduser 0.09 chldsystem 0.1 RSS 56.3 MB)
i canceled the erroneous at jobs by hand and again properly retired the host:
anarcat@angela:fabric-tasks$ ./retire -H cupani.torproject.org retire-all --parent-host=fsn-node-01.torproject.orgstarting tasks at 2024-04-26 05:19:50.855775+00:00checking for ganeti master on host fsn-node-01.torproject.orgganeti node detected with master fsn-node-01.torproject.orgchecking on fsn-node-01.torproject.org if instance cupani.torproject.org is runninginstance cupani.torproject.org not running, no stop requiredscheduling cupani.torproject.org instance removal on host fsn-node-01.torproject.orgscheduling gnt-instance remove --force cupani.torproject.org to run on fsn-node-01.torproject.org in 90 dayswarning: commands will be executed using /bin/shjob 34 at Thu Jul 25 05:19:00 2024scheduling cupani.torproject.org backup disks removal on host bungei.torproject.org and director bacula-director-01.torproject.orgchecking for path "/srv/backups/bacula/cupani.torproject.org/" on bungei.torproject.orgscheduling rm -rf "/srv/backups/bacula/cupani.torproject.org/" to run on bungei.torproject.org in 365 dayswarning: commands will be executed using /bin/shjob 123 at Sat Apr 26 05:20:00 2025checking for path "/srv/backups/pg/cupani/" on bungei.torproject.orgpath /srv/backups/pg/cupani/ not found: [Errno 2] No such filescheduling echo delete client=cupani.torproject.org-fd yes | bconsole to run on bacula-director-01.torproject.org in 365 dayswarning: commands will be executed using /bin/shjob 66 at Sat Apr 26 05:20:00 2025Notice: Revoked certificate with serial 30Notice: Revoked certificate with serial 103cupani.torproject.orgSubmitted 'deactivate node' for cupani.torproject.org with UUID 7a26d96f-98a8-4b80-9468-9b01d791f217completed tasks, elasped: 0:00:20.790689 (user 5.97 system 0.14 chlduser 0.08 chldsystem 0.12 RSS 56.2 MB)
i re-read the RFC again, and the procedure is to retire the backups 12 months after the VM, so that's not 365 days, it's 455 (one year and 3 months) days from now. so, again, canceled by hand and reissued:
anarcat@angela:fabric-tasks$ ./retire -H cupani.torproject.org remove-backups starting tasks at 2024-04-26 05:25:07.781589+00:00checking for path "/srv/backups/bacula/cupani.torproject.org/" on bungei.torproject.orgscheduling rm -rf "/srv/backups/bacula/cupani.torproject.org/" to run on bungei.torproject.org in 455 dayswarning: commands will be executed using /bin/shjob 124 at Fri Jul 25 05:25:00 2025checking for path "/srv/backups/pg/cupani/" on bungei.torproject.orgpath /srv/backups/pg/cupani/ not found: [Errno 2] No such filescheduling echo delete client=cupani.torproject.org-fd yes | bconsole to run on bacula-director-01.torproject.org in 455 dayswarning: commands will be executed using /bin/shjob 67 at Fri Jul 25 05:25:00 2025completed tasks, elasped: 0:00:07.918059 (user 1.75 system 0.07 chlduser 0.05 chldsystem 0.06 RSS 55.2 MB)