Hi Finn Arne,
On Tue, Feb 12, 2013 at 11:07:55AM +0100, Finn Arne Gangstad wrote:
I have a server running v1.5-dev17-17-gcf181c9 which reliably (after
a few hours) gets into a state where it spends 100% in epoll. It still
works as it should, but calls epoll all the time. I'm pretty sure it
On Tue, Feb 12, 2013 at 11:16:32AM +0100, Willy Tarreau wrote:
Just to confirm it doesn't come from commit 1c07b0755, could you please
try to apply the attached patch which reverts it ?
As usual I forgot to attach the patch. Here it is.
Willy
diff --git a/src/ev_epoll.c b/src/ev_epoll.c
index
New haproxy running, may take some hours before it happens again. I'll
see if I can get it nailed down to the second it happens. In the
meantime, I did some digging into the CPU logs, the last incident
happened at 00:02:00 +- 15 seconds, here is the log output from
haproxy around that time
On Tue, Feb 12, 2013 at 12:10:00PM +0100, Finn Arne Gangstad wrote:
New haproxy running, may take some hours before it happens again. I'll
see if I can get it nailed down to the second it happens. In the
meantime, I did some digging into the CPU logs, the last incident
happened at 00:02:00 +-
Monitored cpu usage per second now:
11:52:38 AM 28508223.001.000.004.00 4 haproxy
11:52:39 AM 2850822 43.00 58.000.00 101.00 4 haproxy
11:52:40 AM 2850822 42.00 57.000.00 99.00 4 haproxy
At 11:52:39 the following happened:
Just a quick update on the subject for people following this thread.
Today with Finn Arne's help I could finally spot the issue which is
triggered by the use of observe layer7. This affects all 1.5-dev
versions past 1.5-dev12. After being tested the whole afternoon, the
fix was pushed upstream as
HAProxy community is great
On Tue, Feb 12, 2013 at 8:29 PM, Willy Tarreau w...@1wt.eu wrote:
Just a quick update on the subject for people following this thread.
Today with Finn Arne's help I could finally spot the issue which is
triggered by the use of observe layer7. This affects all
7 matches
Mail list logo
