 |
I had the problem when running on Bareos 20.1. Upgrading to Bareos 21 and I still have the problem. |
|
|
Just ran into this same problem :( Linux Version 21.1.3 |
|
|
Still an issue on 21.1.4 |
|
|
Thanks for your report, after checking the code, we maybe have found the issue. There's a work in progress actually, and I would be able to propose packages to tests soon. Can someone will be able to test them ? (they will be based on master) |
|
|
I've now published normally fixed packages at https://download.bareos.org/bareos/experimental/CD/PR-1265/ We would really appreciate if any of you (or all) can confirm that works for your use case. If confirmed quite quickly, it can and should be part of 21.1.5 Thanks |
|
|
What component do we need to update to test? Director? Client? |
|
|
The patch included concern only the director. Actually master 22.0~pre) is totally compatible with 21 database schema. So you can run the test director without trouble on a 21 database. Doing a dump of your db, and configuration before is always a good thing. |
|
|
The best binaries to test will be those published on 22 September > 17h30 The fix has been improved to handle also just created jobs but that would have not yet started. |
|
|
I loaded the packages in PR-1265 earlier will report back. Not sure how to grab the packages published on >17h30 (?) |
|
|
Same here, loaded the new packages on my director. It will take a couple of days to make sure there is no duplicate. |
|
|
I'm still getting some jobs killed by 25-Sep 11:38 myth-dir JobId 66664: Fatal error: JobId 66658 already running. Duplicate job not allowed. Which I would expect this setting to fix. Now before I would get several of these a day but now I get 1-2. so it's better but still kicking off extras. |
|
|
I'm trying to get a new build out, but we face some trouble with our Jenkins for the moment. would you mind in the meantime to tell be a bit more about your 25-Sep 11:38 myth-dir JobId 66664: Fatal error: JobId 66658 already running. Duplicate job not allowed. What kind of job are they, did they run a pre-post script, what is their status before being run etc. |
|
|
I'm still seeing duplicate jobs : 10965 Backup UltraMagnus ultramagnus Backup Incremental 2022-09-28 20:40:36 2022-09-28 20:40:36 00:00:00 0 0.00 B 1 Canceled 10964 Backup UltraMagnus ultramagnus Backup Incremental 2022-09-28 20:40:19 2022-09-28 20:40:54 00:00:35 1047 664.75 MB 0 Success The second job gets cancelled because I use the following option Allow Duplicate Jobs = no The configuration for the backup is as follows Client { Name = ultramagnus Address = 10.0.1.2 # Symbolic IP Password = XYZ Connection From Client To Director = yes Connection From Director To Client = no Heartbeat Interval = 1 min } Job { Name = "Backup UltraMagnus" Client = "ultramagnus" JobDefs = "DefaultJobLaptop" FileSet = "MacHome" RunOnIncomingConnectInterval = 6h Allow Duplicate Jobs = no Run Before Job = "curl -m 10 --retry 5 http://hc-ping.com/72f87410-63c1-4e56-8002-e06638b67de8/start" Run After Job = "curl -m 10 --retry 5 http://hc-ping.com/72f87410-63c1-4e56-8002-e06638b67de8" Run After Failed Job = "curl -m 10 --retry 5 http://hc-ping.com/72f87410-63c1-4e56-8002-e06638b67de8/fail" } The curl calls are used for monitoring (cronitor) and return instantly. |
|
|
Thanks all for participating, Finally I was able to get the new build out. The new packages are available at https://download.bareos.org/bareos/experimental/CD/PR-1265/ dated of 2022-09-29 15:39 Would you mind to test them as before, and report any failure left. |
|
|
github PR 1265 |
|
|
Will do loaded it on today. Will see what happens. |
|
|
Been running a few days working flawlessly so far. |
|
|
Thanks for the success report, I will do now the rest of the administrative work to make in included into master, and maybe backported to 21 (but no warranties for that to be accepted) |
|
|
PR#1265 merged in master. No backport to previous version has been set. |
- Anonymous
