Hi, I'm seeing regular postscreen segfaults on a test server with minimal traffic. The patterns I noticed from the logs is that it seems to happen when the server gets 2 ~simultaneous connections from the same host:
2024-02-04T14:33:31.876390 info postfix starting the Postfix mail system 2024-02-04T14:33:31.882441 info postfix daemon started -- version 3.8.4, configuration /etc/postfix 2024-02-04T14:34:02.888310 info postfix cache lmdb:/var/lib/postfix/postscreen_cache full cleanup: retained=0 dropped=0 entries 2024-02-04T14:34:02.890666 info postfix CONNECT from [91.239.130.97]:42190 to [172.16.X.Y]:25 2024-02-04T14:34:02.890879 info postfix CONNECT from [91.239.130.97]:42196 to [172.16.X.Y]:25 2024-02-04T14:34:02.890966 info postfix HANGUP after 0 from [91.239.130.97]:42190 in tests before SMTP handshake 2024-02-04T14:34:02.890989 info postfix DISCONNECT [91.239.130.97]:42190 2024-02-04T14:34:02.928708 warning postfix warning: process /usr/libexec/postfix/postscreen pid 93 killed by signal 11 (There's usually no HANGUP/DISCONNECT before the crash, not sure what happened here but was surprised to see the crash as Postfix had just started.) These connections are from an SMTP probe that goes EHLO STARTTLS EHLO QUIT I've not run postscreen previously, so I cannot tell whether this is new, but some testing shows it happens for the following versions: - 3.8.4 on alpine 3.19.0 - 3.8.5 on alpine 3.19.1 but apparently not for 3.8.3 on alpine 3.18.3 I have these postscreen settings: postscreen_dnsbl_sites = zen.spamhaus.org=127.0.0.[2..11] postscreen_dnsbl_action = enforce
_______________________________________________ Postfix-users mailing list -- postfix-users@postfix.org To unsubscribe send an email to postfix-users-le...@postfix.org
