On Mon 13 Jun 2022 at 12:38:37 (+0100), Brad Rogers wrote: > On Mon, 13 Jun 2022 07:24:07 -0400 Greg Wooledge wrote: > >On Mon, Jun 13, 2022 at 04:58:44AM +0000, Russell L. Harris wrote: > >> I suppose I should have shut down some of processes before running > >> Synaptic. > >That shouldn't be necessary.
While I would generally concur, I would not advise, for example, performing the monthly firefox upgrade while the browser is running. > Indeed. It seems more likely that the problems arose because Synaptic > was forcefully terminated. Of course, it may have been necessary due to > a (now) unknowable problem. > > It would be helpful (maybe) to know how long Russell waited before > performing the reset after Synaptic hung. Too late now, but it might > simply have been performing complex 'bookkeeping' tasks. For example, > occasionally, I've seen initrd.img being written out more than once > during an update (although not recently). That can take time. Over the weekend, both ntfs-3g and linux-image-5.10.0 were upgraded, though I split them because I happened to upgrade ntfs-3g just after midnight Friday. A weekly upgrade might upgrade both at the same time. I would also point out that a pair of upgrades like this would be split if you normally upgrade (in apt-get's parlance), but then dist-upgrade because you see a new kernel image being held back. But AFAICT, apt-get does not appear to try to avoid generating initrd twice in the same step. For example, in April, both apparmor and linux-image-5.10.0 caused initrd regeneration on this PC in one step. Not being a synaptic user, I don't know whether it informs users of what it is doing (other than through the logs), but I would have thought it ought to. Cheers, David.
