since F38 it happens to me that the dir for the journals is filled up with a lot of small journals (~5 MB) and something regarding time
Is this still the case with F39 *new* installs (explicit *NOT* upgrades, e.g. F38 => F39)
On 9/27/23 17:44, old sixpack13 wrote:
since F38 it happens to me that the dir for the journals is filled up with a lot of small journals (~5 MB) and something regarding time
I don't know what is the exact journal message you're seeing, but [0] (and [1]) might be related.
[0] https://github.com/systemd/systemd/issues/28250 [1] https://github.com/systemd/systemd/pull/29017
On 9/27/23 17:44, old sixpack13 wrote:
I don't know what is the exact journal message you're seeing, but [0] (and [1]) might be related.
[0] https://github.com/systemd/systemd/issues/28250 [1] https://github.com/systemd/systemd/pull/29017
Exact !
remaining question: is it in new installed F39 too ?
On 9/28/23 01:11, old sixpack13 wrote:
On 9/27/23 17:44, old sixpack13 wrote:
I don't know what is the exact journal message you're seeing, but [0] (and [1]) might be related.
[0] https://github.com/systemd/systemd/issues/28250 [1] https://github.com/systemd/systemd/pull/29017
Exact !
remaining question: is it in new installed F39 too ?
As far as I know yes, the potential fixes from [1] are not in any released or stable systemd version.
On Thu, 2023-09-28 at 09:28 +0200, František Šumšal wrote:
On 9/28/23 01:11, old sixpack13 wrote:
On 9/27/23 17:44, old sixpack13 wrote:
I don't know what is the exact journal message you're seeing, but [0] (and [1]) might be related.
[0] https://github.com/systemd/systemd/issues/28250 [1] https://github.com/systemd/systemd/pull/29017
Exact !
remaining question: is it in new installed F39 too ?
As far as I know yes, the potential fixes from [1] are not in any released or stable systemd version.
It looks like they are in 254.5, which is on its way to updates-testing for F39 now: https://bodhi.fedoraproject.org/updates/FEDORA-2023-3355cdb0a8
On Thu, 2023-09-28 at 09:28 +0200, František Šumšal wrote:
It looks like they are in 254.5, which is on its way to updates-testing for F39 now: https://bodhi.fedoraproject.org/updates/FEDORA-2023-3355cdb0a8
unfortunately, nit fixed with 254.5-2.fc39
still seeing: - journal: Realtime clock jumped backwards relative to last journal entry, rotating - Journal file uses a different sequence number ID, rotating
On Sat, 30 Sep 2023 17:50:31 -0000 "old sixpack13" sixpack13@online.de wrote:
On Thu, 2023-09-28 at 09:28 +0200, František Šumšal wrote:
It looks like they are in 254.5, which is on its way to updates-testing for F39 now: https://bodhi.fedoraproject.org/updates/FEDORA-2023-3355cdb0a8
unfortunately, nit fixed with 254.5-2.fc39
still seeing:
- journal: Realtime clock jumped backwards relative to last journal
entry, rotating
- Journal file uses a different sequence number ID, rotating
This is just a wild guess, but the tzdata package was recently dropped as required by some packages. Is it possible some fallout from this is the source of your problem? Do you have the tzdata package(s) installed?
Another remote possibility is that your system isn't maintaining time properly. Could the little pancake CR2032 battery be dead?
On Sat, 30 Sep 2023 17:50:31 -0000 "old sixpack13" <sixpack13(a)online.de> wrote:
This is just a wild guess, but the tzdata package was recently dropped as required by some packages. Is it possible some fallout from this is the source of your problem? Do you have the tzdata package(s) installed?
rpm -qa|grep tzdata tzdata-2023c-2.fc39.noarch tzdata-java-2023c-2.fc39.noarch
more needed ? box was a new install of F38 with the updated ISO at 2023-07-28 and upgraded to F39 since beta is out
Another remote possibility is that your system isn't maintaining time properly. Could the little pancake CR2032 battery be dead?
the box is new build from: 2021-10-12 my experiences with CMOS Batteries are: exchange needed when elder > 3 years
On Mon, 02 Oct 2023 17:03:12 -0000 "old sixpack13" sixpack13@online.de wrote:
On Sat, 30 Sep 2023 17:50:31 -0000 "old sixpack13" <sixpack13(a)online.de> wrote:
This is just a wild guess, but the tzdata package was recently dropped as required by some packages. Is it possible some fallout from this is the source of your problem? Do you have the tzdata package(s) installed?
rpm -qa|grep tzdata tzdata-2023c-2.fc39.noarch tzdata-java-2023c-2.fc39.noarch more needed ?
No, that would do it.
Another remote possibility is that your system isn't maintaining time properly. Could the little pancake CR2032 battery be dead?
the box is new build from: 2021-10-12 my experiences with CMOS Batteries are: exchange needed when elder > 3 years
I'd give this a pass as well. My thought was that the systemd messages from boot for the journal, before chronyd was loaded, were off because the time was different than the time that chronyd pulled from the web.
Given your data, I don't think either of these is your issue. I'm out of ideas.
-
Adam Williamson -
František Šumšal -
old sixpack13 -
stan