Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.user > #222443
| From | Gene Heskett <gheskett@shentel.net> |
|---|---|
| Newsgroups | linux.debian.user |
| Subject | Re: Question on Unattended Upgrades |
| Date | 2020-05-13 06:50 +0200 |
| Message-ID | <A5Rvc-2ef-3@gated-at.bofh.it> (permalink) |
| References | <A5EHD-2qM-1@gated-at.bofh.it> <A5PWp-1hU-1@gated-at.bofh.it> |
| Organization | none,nada,zip |
On Tuesday 12 May 2020 22:49:48 Tom Dial wrote: > On 5/12/20 09:05, Gene Heskett wrote: > > On Tuesday 12 May 2020 10:30:04 Celejar wrote: > >> On Tue, 12 May 2020 06:03:52 -0400 > >> > >> Gene Heskett <gheskett@shentel.net> wrote: > >>> On Tuesday 12 May 2020 05:07:04 l0f4r0@tuta.io wrote: > >>>> Hi, > >>>> > >>>> 12 mai 2020 à 08:22 de keifer.bly@gmail.com: > >>>>> Is there a way to configure it to automatically restart when a > >>>>> package that needs to be restarted is upgraded? > >>>> > >>>> I think 'Unattended-Upgrade::Automatic-Reboot "true";' should do > >>>> it in /etc/apt/apt.conf.d/50unattended-upgrades. If I were you, I > >>>> would check /etc/apt/apt.conf.d/20auto-upgrades as well because > >>>> those are the 2 most important configuration files for > >>>> unattended-upgrades. > >>>> > >>>> Best regards, > >>>> l0f4r0 > >>> > >>> I can see a major disaster in the making, what if that reboot was > >>> commanded in the middle of your nightly backup run, leaving your > >>> backup program with a totally bogus database it cannot recover > >>> from? > >> > >> That would be unfortunate - but surely a serious backup program > >> would be designed to be robust enough that ending up "with a > >> totally bogus database it cannot recover from" should never happen? > >> What if you have some sort of system crash during a backup? Do you > >> have a specific backup system in mind? > > > > You are trying to justify an automatic reboot, very bad idea. > > Granted, most such would be ok, and will leave a requester popup to > > advise the admin, doing its normal routine in the meantime. > > > > I happen to use amanda to backup 5 machines here, but any backup > > suite that maintains a recovery database is going to be in deep > > doodoo if it gets rebooted in the middle of a backup. I probably do > > better than most but a worst case scenario here if everything in > > place works, would only lose me that days data because I do keep > > separate copies of that data. But 60 days of that is 33GB, the > > biggest single entry in my nearly 70 items long list. > > I don't know about Amanda other than understanding it to be a > commercial quality backup system. In my working days as an > administrator, my organization used HP Data Protector (now Micro > Focus). As I recall, it had a proprietary transactional database that > should have been up to date as of the last successfully completed file > backups before a crash. It also, at least in our usage, backed up its > database as the first action in a cycle, so in the worst case, > reloading the database and rerunning the faulted backup would restore > synchronization of the database and data. And if all else failed, I > think there were procedures for rebuilding the database from the > tapes. I wonder if Amanda does not provide similar features. > > We also did not do automatic updates, there being organizational > requirements to test configuration changes before installing them on > production systems. > > I don't consider automatic updates a very good idea, although out of > laziness in retirement I allow automatic security updates. Unplanned > reboots seem an even worse idea inasmuch as they occasionally will > either fail or affect system function. (Updating without rebooting > also may do that, of course). > > >>> Frankly the update shouldn't be allowed if your backup is actually > >>> running. So it might be safer to schedule the update and reboot > >>> if needed before the backup starts. That means one cron driven > >>> script does it all in the sequence desired. > > With respect, I would run the backup first and schedule upgrades and > possible reboots to follow, and perhaps depend on success of, the > backup. > > Regards, > Tom Dial > We are in violent agreement on that. I've never done an "automatic" upgrade, although apt does do a update to keep its database as to whats available automaticly. The only fault with the list --upgradeable output is that it does not filter out what is pinned, which on the other 4 machines here, is the kernel and its libs because its a special realtime built kernel for running cnc machinery. So they don't often get rebooted unless something has obviously screwed the moose. X related stuff for instance often qualifies. Although x development's pace has slowed, the gui's it makes possible hasn't. As far as amanda's database is concerned, until the backup has been done, that database is effectively yesterdays, so I wrote a wrapper that appends that now up to date database to the just completed backup so the database in THAT backup is up to date. Gives me a little more peace of mind. Thanks, stay well and safe now. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis Genes Web page <http://geneslinuxbox.net:6309/gene>
Back to linux.debian.user | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
Re: Re: Question on Unattended Upgrades Victor Sudakov <vas@sibptus.ru> - 2020-05-12 07:00 +0200
Re: Re: Question on Unattended Upgrades Keifer Bly <keifer.bly@gmail.com> - 2020-05-12 07:30 +0200
Re: Question on Unattended Upgrades Marco Möller <talby@debianlists.mobilxpress.net> - 2020-05-13 15:10 +0200
Re: Question on Unattended Upgrades Charles Curley <charlescurley@charlescurley.com> - 2020-05-12 08:20 +0200
Re: Question on Unattended Upgrades Keifer Bly <keifer.bly@gmail.com> - 2020-05-12 08:30 +0200
Re: Question on Unattended Upgrades l0f4r0@tuta.io - 2020-05-12 11:10 +0200
Re: Question on Unattended Upgrades Gene Heskett <gheskett@shentel.net> - 2020-05-12 12:10 +0200
Re: Question on Unattended Upgrades l0f4r0@tuta.io - 2020-05-12 12:20 +0200
Re: Question on Unattended Upgrades Celejar <celejar@gmail.com> - 2020-05-12 16:40 +0200
Re: Question on Unattended Upgrades Gene Heskett <gheskett@shentel.net> - 2020-05-12 17:10 +0200
Re: Question on Unattended Upgrades Celejar <celejar@gmail.com> - 2020-05-12 17:20 +0200
Re: Question on Unattended Upgrades Tom Dial <tddial@comcast.net> - 2020-05-13 05:10 +0200
Re: Question on Unattended Upgrades Keifer Bly <keifer.bly@gmail.com> - 2020-05-13 06:40 +0200
Re: Question on Unattended Upgrades l0f4r0@tuta.io - 2020-05-13 10:00 +0200
Re: Question on Unattended Upgrades Gene Heskett <gheskett@shentel.net> - 2020-05-13 10:20 +0200
Re: Question on Unattended Upgrades Gene Heskett <gheskett@shentel.net> - 2020-05-13 06:50 +0200
Re: Question on Unattended Upgrades Keifer Bly <keifer.bly@gmail.com> - 2020-05-13 03:20 +0200
Re: Question on Unattended Upgrades rhkramer@gmail.com - 2020-05-12 13:50 +0200
Re: Question on Unattended Upgrades Greg Wooledge <wooledg@eeg.ccf.org> - 2020-05-12 14:00 +0200
Re: Question on Unattended Upgrades Reco <recoverym4n@enotuniq.net> - 2020-05-12 14:00 +0200
Re: Question on Unattended Upgrades David Wright <deblis@lionunicorn.co.uk> - 2020-05-12 18:00 +0200
Re: Question on Unattended Upgrades John Covici <covici@ccs.covici.com> - 2020-05-12 21:30 +0200
Re: Question on Unattended Upgrades David Wright <deblis@lionunicorn.co.uk> - 2020-05-12 22:50 +0200
Re: Question on Unattended Upgrades John Covici <covici@ccs.covici.com> - 2020-05-12 23:50 +0200
Re: Question on Unattended Upgrades l0f4r0@tuta.io - 2020-05-13 00:20 +0200
csiph-web