I have a Tumbleweed system that was last updated Oct 2019. I am running Trac (https://trac.edgewall.org/), which has been slow to move to Python 3 - and it is just about there now. I did not want to chance messing up the Python 2 setup that has been working fine. So, I abstained from updating until the dust settled. Given the amount of time that has passed, is it reasonable to expect that an update in the near future would work? I worry about things like the /usr/lib64 move. And probably other things that have been rather big. All done at once. Has anyone else left such a big time between Tumbleweed updates? I do not generally do that with my other Tumbleweed systems. This was a special case. I am using this system in production. So I know, do a backup. Everything of interest is on a separate (from the OS) RAID volume. I might disconnect that during the update. I have generally been lucky. I am running snapper. Aside from a couple small installs of some packages, thee main snapshot is from Oct 2019. I do not want it to be removed. It is rather old. I guess I need to check the rules for this to be sure it won't be removed as part of a cleanup. Anything else to think of? -- Roger Oberholtzer
Roger Oberholtzer wrote:
Given the amount of time that has passed, is it reasonable to expect that an update in the near future would work? I worry about things like the /usr/lib64 move. And probably other things that have been rather big. All done at once. Has anyone else left such a big time between Tumbleweed updates?
No, not quite that long, but I do have two systems that I update maybe once a year, usually a little longer. Sofar it has worked fine. FWIW, I'm just now running a 'dup' on one of them, some 3000 updates. I thought the big change was the usr merge, but I don't really keep up with TW. -- Per Jessen, Zürich (16.6°C)
Per Jessen wrote:
No, not quite that long, but I do have two systems that I update maybe once a year, usually a little longer. Sofar it has worked fine. FWIW, I'm just now running a 'dup' on one of them, some 3000 updates.
Just a brief update - the 'dup' completed, but the system did not boot, unable to mount root. I'm debating whether to hook up a serial console or boot an installer to see if I can spot the issue. -- Per Jessen, Zürich (0.0°C)
Per Jessen wrote:
Per Jessen wrote:
No, not quite that long, but I do have two systems that I update maybe once a year, usually a little longer. Sofar it has worked fine. FWIW, I'm just now running a 'dup' on one of them, some 3000 updates.
Just a brief update - the 'dup' completed, but the system did not boot, unable to mount root. I'm debating whether to hook up a serial console or boot an installer to see if I can spot the issue.
I booted the tw installer over pxe, I was able to mount the root filesystem (after manually loading the blacklisted filesystem module). I decided it was worth rebuilding the initrd, which did solve the problem. I diff'ed the two initrds, but I don't see any immediate cause for the problem. -- Per Jessen, Zürich (19.8°C) http://www.dns24.ch/ - free dynamic DNS, made in Switzerland.
Per do you know much about Sun Microsystems ; perhaps rc.d implentation ? On Apr-13-2022 06:29AM, Per Jessen wrote:
Per Jessen wrote:
Per Jessen wrote:
No, not quite that long, but I do have two systems that I update maybe once a year, usually a little longer. Sofar it has worked fine. FWIW, I'm just now running a 'dup' on one of them, some 3000 updates. Just a brief update - the 'dup' completed, but the system did not boot, unable to mount root. I'm debating whether to hook up a serial console or boot an installer to see if I can spot the issue. I booted the tw installer over pxe, I was able to mount the root filesystem (after manually loading the blacklisted filesystem module). I decided it was worth rebuilding the initrd, which did solve the problem. I diff'ed the two initrds, but I don't see any immediate cause for the problem.
13.04.22 09:19 - Roger Oberholtzer:
I have a Tumbleweed system that was last updated Oct 2019. I am running Trac (https://trac.edgewall.org/), which has been slow to move to Python 3 - and it is just about there now. I did not want to chance messing up the Python 2 setup that has been working fine. So, I abstained from updating until the dust settled.
Given the amount of time that has passed, is it reasonable to expect that an update in the near future would work? I worry about things like the /usr/lib64 move. And probably other things that have been rather big. All done at once. Has anyone else left such a big time between Tumbleweed updates? I do not generally do that with my other Tumbleweed systems. This was a special case. I am using this system in production. So I know, do a backup. Everything of interest is on a separate (from the OS) RAID volume. I might disconnect that during the update.
I have generally been lucky. I am running snapper. Aside from a couple small installs of some packages, thee main snapshot is from Oct 2019. I do not want it to be removed. It is rather old. I guess I need to check the rules for this to be sure it won't be removed as part of a cleanup. Anything else to think of?
I have a Tumbleweed system which was installed in 2019 and regularly updated since then. Due to all the recent changes (usr-merge, ...) I thought it might be a good idea to do a fresh install (on a spare partition) using a recent snapshot. Using that freshly installed system i "discovered": That freshly installed Tumbleweed system did not create users starting with UID=1000=NAME GID=100=users but with UID=1000=NAME GID=1000=NAME (like UBUNTU) I had not heard of that before (and my since 2019 regularly updated Tumbleweed still does not behave like this). Before considering a fresh install with a recent snapshot it might be worth to be aware of that change. Regards Hagen
On Thu, Apr 14, 2022 at 3:38 PM Hagen Buliwyf
That freshly installed Tumbleweed system did not create users starting with
UID=1000=NAME GID=100=users
but with
UID=1000=NAME GID=1000=NAME (like UBUNTU)
I had not heard of that before
There was a rather long discussion on the factory list before this change ...
participants (5)
-
-pj
-
Andrei Borzenkov
-
Hagen Buliwyf
-
Per Jessen
-
Roger Oberholtzer