You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi
I have used c-icap for a long time and I have recently encountered a problem: c-icap got stuck while shutting down the process.
As I could figure out it happens in really rare conditions: there are two subsequent commands "reconfigure" and "stop" sent to command socket with small pause between them (I guess 0.3 - 0.5s):
As I understand c-icap get stuck while it is waiting for child exit that have to be killed after reconfiguration command. Probably, the cause is in this condition in wait_childs_to_exit function (mpmt_server.c):
if (q->childs[i].to_be_killed != IMMEDIATELY) {
ci_debug_printf(5, "Child %d not signaled yet!\n", q->childs[i].pid);
continue;
}
to_be_killed here has GRACEFULLY value for those child processes.
Could you please take a look at this issue?
I understand that scenario is quite weird but it happened in my project. And I can't restart the whole service again due to this stuck (process still exists).
The text was updated successfully, but these errors were encountered:
ae-ignatiev
changed the title
c-icap get stuck while shutting down (reconfigure command then stop command)
c-icap gets stuck while shutting down (reconfigure command then stop command)
Jun 30, 2023
Hi
I have used c-icap for a long time and I have recently encountered a problem: c-icap got stuck while shutting down the process.
As I could figure out it happens in really rare conditions: there are two subsequent commands "reconfigure" and "stop" sent to command socket with small pause between them (I guess 0.3 - 0.5s):
The reproducing is quite stable.
As I understand c-icap get stuck while it is waiting for child exit that have to be killed after reconfiguration command. Probably, the cause is in this condition in wait_childs_to_exit function (mpmt_server.c):
to_be_killed
here has GRACEFULLY value for those child processes.Could you please take a look at this issue?
I understand that scenario is quite weird but it happened in my project. And I can't restart the whole service again due to this stuck (process still exists).
Processes/threads configuration in config-file:
Version: 0.5.10
CentOS 7.9
Thank you!
The text was updated successfully, but these errors were encountered: