henrik.nordstrom@ida.his.se writes:
>> Background Restore takes too long if there are no events occuring
>> (50 objects read once a second), added a flag to cause squid to block
>> until the restore is complete.
>
>Not quite true. It restores 50 objects, then handles any pending
>network data, restores 50 more objects and so on (no delay). The
>more concurrent activity, the slower it reloads.
I think Mark is referring to a bug which I fixed between beta10 and
beta11. comm_select() used a one second timeout and it would block if
there was no pending I/O, even during the rebuild.
@@ -692,7 +699,7 @@
if (nfds == 0)
return COMM_SHUTDOWN;
while (1) {
- poll_time.tv_sec = 1;
+ poll_time.tv_sec = sec > 1 ? 1 : 0;
poll_time.tv_usec = 0;
num = select(maxfd, &readfds, &writefds, &exceptfds, &poll_time);
if (num >= 0)
Duane W.
Received on Thu Jun 13 1996 - 13:06:51 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:32:30 MST