The patch is visually fine, and I was actually successfull in testing
pre/post behaviour. As in: client with multiple <connection> blocks,
and using the management interface to make it connect to server A
("push ping, ping-restart") first, then server B ("do not push anything
ping-related"). Without the patch, the client would die with --ping-restart
timeout, with the patch it happily does "no ping restart".
It needs SIGUSR1 restarts to trigger the issue. SIGHUP restarts do a
"full config reload" and never display "lingering pushed config" problems.
Testing has uncovered more save & restore bugs, related to compression
and "--route-gateway". At least...
Your patch has been applied to the master and release/2.5 branch (bugfix),
but I'm not sure I want to really backport it to 2.4 - it's not a crucial
crashbug, not easy to trigger, and if you're annoyed by it, just upgrade.
I have not tested the 2.5 backport in full depth, just a basic t_client
run (but the change is straightforward enough, and it applied without
much friction).
commit 5a2ed714d14acb2312d27fe40e300be96d970c27 (master)
commit a0e844c892d6f67977bf8e9162cbc43a3f64ba46 (release/2.5)
Author: Arne Schwabe
Date: Wed Mar 17 17:00:38 2021 +0100
Restore also ping related options on a reconnect
Signed-off-by: Arne Schwabe <[email protected]>
Acked-by: Antonio Quartulli <[email protected]>
Message-Id: <[email protected]>
URL:
https://www.mail-archive.com/[email protected]/msg21675.html
Signed-off-by: Gert Doering <[email protected]>
--
kind regards,
Gert Doering
_______________________________________________
Openvpn-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openvpn-devel
