It is time for your update of rspamd, let’s go to 2.5. We propose to test an update when the features and the bug fixes are interesting. Here we are
yum update rspamd --enablerepo=nethserver-testing
The changelog
As far I know the settings are not modified by this version.
Please test and report.
Got some errors after installation but it seems to work so maybe we can ignore it. There’s already an open issue.
Sep 10 17:16:49 server2 rspamd[27784]: <xq3dqm>; cfg; rspamd_language_detector_init: cannot compile stop words for 0 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 7; pattern: a\x{10d}koli
Sep 10 17:16:49 server2 rspamd[27784]: <xq3dqm>; cfg; rspamd_language_detector_init: cannot compile stop words for 1 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{441}\x{430}\x{43c}
Sep 10 17:16:49 server2 rspamd[27784]: <xq3dqm>; cfg; rspamd_language_detector_init: cannot compile stop words for 2 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{906}\x{92b}\x{942}
Sep 10 17:16:49 server2 rspamd[27784]: <xq3dqm>; cfg; rspamd_language_detector_init: cannot compile stop words for 3 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{647}\x{646}\x{627}\x{644}\x{643}
Thanks for your work!
@mrmarkuz is it something that you can reproduce or only at the first start of the rspamd after the upgrade ?
I do not have this in my maillog 
I think I can reproduce it
I just updated from rspamd 2.4 and saw the errors in rspamd history UI, I did not test more yet…
I think it is related to the processor you used with your KVM/vmware
With the default (I use proxmox)
Sep 11 17:47:33 ns7dev13 rspamd[25095]: <>; ; rspamd_regexp_library_init: pcre is compiled with JIT for x86 64bit (little endian + un
aligned)
Sep 11 17:47:33 ns7dev13 rspamd[25095]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 0 language group:
regexp parsing error: 'character value in \x{...} sequence is too large' at position 7; pattern: a\x{10d}koli
Sep 11 17:47:33 ns7dev13 rspamd[25095]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 1 language group:
regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{441}\x{430}\x{43c}
Sep 11 17:47:33 ns7dev13 rspamd[25095]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 2 language group:
regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{906}\x{92b}\x{942}
Sep 11 17:47:33 ns7dev13 rspamd[25095]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 3 language group:
regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{647}\x{646}\x{627}\x{644}\x{643}
Sep 11 17:47:33 ns7dev13 rspamd[25095]: <dg4msr>; cfg; rspamd_language_detector_init: loaded 46 languages, 33122 trigramms
I can reproduce but if I change the processor to the host one it disappears
Sep 11 18:07:02 ns7dev13 rspamd[2073]: <>; ; rspamd_regexp_library_init: pcre is compiled with JIT for x86 64bit (little endian + una
ligned)
Sep 11 18:07:02 ns7dev13 rspamd[2073]: <dg4msr>; cfg; rspamd_language_detector_init: loaded 46 languages, 33122 trigrammsConfirmed when I go back to default I have back the error message
11994:Sep 11 18:27:05 ns7dev13 rspamd[2558]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 0 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 7; pattern: a\x{10d}koli
11995:Sep 11 18:27:05 ns7dev13 rspamd[2558]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 1 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{441}\x{430}\x{43c}
11996:Sep 11 18:27:05 ns7dev13 rspamd[2558]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 2 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{906}\x{92b}\x{942}
11997:Sep 11 18:27:05 ns7dev13 rspamd[2558]: <dg4msr>; cfg; rspamd_language_detector_init: cannot compile stop words for 3 language group: regexp parsing error: 'character value in \x{...} sequence is too large' at position 6; pattern: \x{647}\x{646}\x{627}\x{644}\x{643}Confirmed. I use proxmox with default(kvm64).
And after setting the processor to host there are no errors anymore:
could you explain why multiple cores are needed to fix that error? how did you come to this solution?
Reading this : https://github.com/rspamd/rspamd/issues/3124#issuecomment-552459148
I guess and tested to change the default KVM CPU, also on my real server I use the host CPU and I stated that the error was not visible on my server.
Since the update to version 2.5, I have a high number of identical errors in the log every morning after system startup as shown in the picture.
Regards…
Uwe
where the screenshot comes , I do not know ?
However the error on the map is not in our hand, we know it is an error that you can see from times to times.
Sometimes an error in your dns server can cause it.
