On Fri, Aug 19, 2022 at 11:37:47PM +0200, Vincent Bernat wrote: > On 2022-08-19 23:09, Ionel GARDAIS wrote: > > Aug 19 22:09:09 haproxy-2 haproxy[1280]: [WARNING] (1280) : Failed to > > connect to the old process socket '/run/haproxy/admin.sock' > > Aug 19 22:09:09 haproxy-2 haproxy[1280]: [ALERT] (1280) : Failed to get > > the sockets from the old process! > > There was a change in 2.6.0 (but not in 2.6.3) where "expose-fd listeners" > for stats socket is not needed anymore. Is the line present in your > configuration? (grep admin.sock /etc/haproxy/haproxy.cfg) > > What's the output of systemctl cat haproxy? > > > Aug 19 22:09:09 haproxy-2 haproxy[1280]: [NOTICE] (1280) : New worker > > (1282) forked > > Aug 19 22:09:09 haproxy-2 haproxy[1280]: [NOTICE] (1280) : Loading > > success. > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: [WARNING] (1282) : Server > > bck-speedtest/go-v6 is DOWN, reason: Layer4 connection problem, info: > > "Connection> > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: Server bck-speedtest/go-v6 is > > DOWN, reason: Layer4 connection problem, info: "Connection refused", check > > dur> > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: Server bck-speedtest/go-v6 is > > DOWN, reason: Layer4 connection problem, info: "Connection refused", check > > dur> > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: [WARNING] (1282) : Server > > bck-nuxeo-arch/nuxeo-arch is DOWN, reason: Layer4 connection problem, info: > > "No r> > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: [ALERT] (1282) : backend > > 'bck-nuxeo-arch' has no server available! > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: Server bck-nuxeo-arch/nuxeo-arch > > is DOWN, reason: Layer4 connection problem, info: "No route to host", check> > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: Server bck-nuxeo-arch/nuxeo-arch > > is DOWN, reason: Layer4 connection problem, info: "No route to host", check> > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: backend bck-nuxeo-arch has no > > server available! > > Aug 19 22:09:09 haproxy-2 haproxy[1282]: backend bck-nuxeo-arch has no > > server available! > > [...] // others few backends not responding > > // then > > Are these expected?
That's what I would like to know as well. I'm bothered by the systemd message indicating that haproxy is not responding, I don't really know what it corresponds to, and could possibly indicate that the sd_notify() message was not sent before switching to wait mode. Ideally the startup messages from 2.6.2 where it's working fine would help as well. Thanks, Willy
