Re: RabbitMQ annoying disconnections [kolla-ansible, rabbitmq]
Hi, I’ve tried to set heartbeat_in_pthread = true, but nothing has changed after kolla-ansible reconfigure - still got disconnections and heartbeat_in_pthread option was present inside configs on all nodes. But in the meantime I tried to alternate rabbit configuration and I realized that Kolla ignored /etc/kolla/config/rabbitmq.conf file (of course at first I checked permissions, etc.) So I tried to put heartbeat = 600 inside /etc/kolla/config/rabbitmq/rabbitmq.conf - this overwrote rabbitmq.conf totally (I mean Kolla didn’t just add heartbeat = 600 to rabbitmq.conf on each controller node, but put ONLY this line in all rabbitmq.conf files ! - rest of the file was cleared). So finally I tried to create separate config for each controller node. So I copied default content of
/etc/kolla/rabbitmq/rabbitmq.conf
on each controller (because of course these files differs from each other - they have different IPs inside) to
/etc/kolla/config/rabbitmq/controller1(2,3)/rabbitmq.conf
And finally heartbeat = 600 started to work - no more disconnections. I left Oslo heartbeats and timeouts intact (with the values I’ve set previously) and I don’t have disconnections from services side too.
So there are two questions:
1. Why heartbeat_in_pthread didn’t change situation? And if it’s deprecated it was kind of risky to use it long term (after upgrade to wallaby it would stop working anyway?) 2. Why Kolla-ansible ignored rabbitmq.conf file if it is located in /etc/kolla/config, and overwrites ALL content of rabbitmq.conf file when it is located in /etc/kolla/config/rabbitmq/ ?
Best regards Adam Tomas
Wiadomość napisana przez Herve Beraud <hberaud@redhat.com mailto:hberaud@redhat.com> w dniu 21.09.2021, o godz. 08:12:
Le sam. 18 sept. 2021 à 16:12, bkslash <bkslash@poczta.onet.pl mailto:bkslash@poczta.onet.pl> a écrit : Hi Ben, Thank you for the answer. I’m using open stack Victoria deployed with Kolla-ansible. So I assume I should add (in /etc/kolla/config/global.conf)
[oslo_messaging_rabbit] heartbeat_in_pthread = True
But for Victoria (nova sample configuration file) it seems to be deprecated? # DEPRECATED: Run the health check heartbeat thread through a native python # thread by default. If this option is equal to False then the health check # heartbeat will inherit the execution model from the parent process. For # example if the parent process has monkey patched the stdlib by using # eventlet/greenlet then the heartbeat will be run through a green thread # (boolean value) # This option is deprecated for removal. # Its value may be silently ignored in the future. #heartbeat_in_pthread = true So will it work?
Yes it will work.
Few months ago I proposed to remove this deprecation and to keep the option available [1], unfortunately my removal isn't yet merged. Only the default value is changed and the option will be turned on by default.
[1] https://review.opendev.org/c/openstack/oslo.messaging/+/800621 https://review.opendev.org/c/openstack/oslo.messaging/+/800621
Best regards Adam Tomas
Wiadomość napisana przez Ben Nemec <openstack@nemebean.com mailto:openstack@nemebean.com> w dniu 17.09.2021, o godz. 18:57:
This sounds like you're hitting the heartbeat/eventlet bug. Basically eventlet won't schedule the heartbeat thread unless something is regularly waking up the api threads. Depending on what release you're running, you may be able to set heartbeat_in_pthread[0] to fix this.
0: https://opendev.org/openstack/oslo.messaging/src/branch/master/oslo_messagin... https://opendev.org/openstack/oslo.messaging/src/branch/master/oslo_messaging/_drivers/impl_rabbit.py#L90
On 9/17/21 7:00 AM, Adam Tomas wrote:
Hi, after some struggling I have almost „clear” logs (clear=error free :) ) Almost…. RabbitMQ keeps disconnecting sessions and there is a huge amount of disconnect errors in all logs (see below). I found this bug description: https://bugzilla.redhat.com/show_bug.cgi?id=1711794 https://bugzilla.redhat.com/show_bug.cgi?id=1711794 <https://bugzilla.redhat.com/show_bug.cgi?id=1711794 https://bugzilla.redhat.com/show_bug.cgi?id=1711794> in which we can read as follows: "this is a recoverable issue that is already handled by how oslo.messaging is designed. disconnection is not an error and should not be reported as such in the logs.” but… It is reported :( And produces tons of logs. I tried to modify heartbeat values - helped a little bit, but I had to increase [database] max_pool_size = 5 and that of course multiplied number of disconnection errors by 5 :( [oslo_messaging_rabbit] heartbeat_timeout_threshold = 720 heartbeat_interval = 360 heartbeat_rate = 4 How to avoid this disconnections?? Or what to do to avoid writing this „errors”to logs? Best regards Adam Tomas /var/log/kolla/rabbitmq/rabbitXXX.log 2021-09-17 06:43:08.771 [error] <0.16056.678> closing AMQP connection <0.16056.678> (x.x.x.x00:44882 -> x.x.x.x00:5672 - mod_wsgi:959:aaa12af5-bb0b-456e-a129-a466eccc24f0): missed heartbeats from client, timeout: 60s 2021-09-17 06:43:08.773 [info] <0.7412.679> Closing all channels from connection 'x.x.x.x00:44882 -> x.x.x.x00:5672' because it has been closed 2021-09-17 06:43:09.282 [info] <0.4574.679> accepting AMQP connection <0.4574.679> (x.x.x.x11:58170 -> x.x.x.x00:5672) 2021-09-17 06:43:09.283 [info] <0.28410.678> accepting AMQP connection <0.28410.678> (x.x.x.x11:58172 -> x.x.x.x00:5672) 2021-09-17 06:43:09.285 [info] <0.28410.678> Connection <0.28410.678> (x.x.x.x11:58172 -> x.x.x.x00:5672) has a client-provided name: magnum-conductor:954:c05a1221-6915-4414-ba93-7db5450ff394 2021-09-17 06:43:09.285 [info] <0.4574.679> Connection <0.4574.679> (x.x.x.x11:58170 -> x.x.x.x00:5672) has a client-provided name: magnum-conductor:967:08f29ce5-d6ee-43ae-87a9-e8c8856f8c93 2021-09-17 06:43:09.286 [info] <0.4574.679> connection <0.4574.679> (x.x.x.x11:58170 -> x.x.x.x00:5672 - magnum-conductor:967:08f29ce5-d6ee-43ae-87a9-e8c8856f8c93): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:09.286 [info] <0.28410.678> connection <0.28410.678> (x.x.x.x11:58172 -> x.x.x.x00:5672 - magnum-conductor:954:c05a1221-6915-4414-ba93-7db5450ff394): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:09.504 [error] <0.14465.678> closing AMQP connection <0.14465.678> (x.x.x.x00:44914 -> x.x.x.x00:5672 - mod_wsgi:964:ced74779-2794-40f6-a01e-276bceb5d7b3): missed heartbeats from client, timeout: 60s 2021-09-17 06:43:09.507 [info] <0.14985.679> Closing all channels from connection 'x.x.x.x00:44914 -> x.x.x.x00:5672' because it has been closed 2021-09-17 06:43:09.521 [info] <0.11574.678> accepting AMQP connection <0.11574.678> (x.x.x.x00:46834 -> x.x.x.x00:5672) 2021-09-17 06:43:09.523 [info] <0.11574.678> Connection <0.11574.678> (x.x.x.x00:46834 -> x.x.x.x00:5672) has a client-provided name: mod_wsgi:964:ced74779-2794-40f6-a01e-276bceb5d7b3 2021-09-17 06:43:09.525 [info] <0.11574.678> connection <0.11574.678> (x.x.x.x00:46834 -> x.x.x.x00:5672 - mod_wsgi:964:ced74779-2794-40f6-a01e-276bceb5d7b3): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:09.654 [info] <0.31575.679> accepting AMQP connection <0.31575.679> (x.x.x.x01:55304 -> x.x.x.x00:5672) 2021-09-17 06:43:09.656 [info] <0.31575.679> Connection <0.31575.679> (x.x.x.x01:55304 -> x.x.x.x00:5672) has a client-provided name: magnum-conductor:7:5a1cc3cc-150d-417a-9652-1279597c3db7 2021-09-17 06:43:09.656 [info] <0.31180.678> accepting AMQP connection <0.31180.678> (x.x.x.x01:55306 -> x.x.x.x00:5672) 2021-09-17 06:43:09.658 [info] <0.31575.679> connection <0.31575.679> (x.x.x.x01:55304 -> x.x.x.x00:5672 - magnum-conductor:7:5a1cc3cc-150d-417a-9652-1279597c3db7): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:09.658 [info] <0.31180.678> Connection <0.31180.678> (x.x.x.x01:55306 -> x.x.x.x00:5672) has a client-provided name: magnum-conductor:954:c8deffa4-7d67-46be-9b9d-09fa20544a2b 2021-09-17 06:43:09.660 [info] <0.2757.679> accepting AMQP connection <0.2757.679> (x.x.x.x11:58174 -> x.x.x.x00:5672) 2021-09-17 06:43:09.660 [info] <0.31180.678> connection <0.31180.678> (x.x.x.x01:55306 -> x.x.x.x00:5672 - magnum-conductor:954:c8deffa4-7d67-46be-9b9d-09fa20544a2b): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:09.662 [info] <0.2757.679> Connection <0.2757.679> (x.x.x.x11:58174 -> x.x.x.x00:5672) has a client-provided name: magnum-conductor:7:cebf81c9-95f6-48fa-8bde-30c01243d315 2021-09-17 06:43:09.663 [info] <0.2757.679> connection <0.2757.679> (x.x.x.x11:58174 -> x.x.x.x00:5672 - magnum-conductor:7:cebf81c9-95f6-48fa-8bde-30c01243d315): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:09.814 [error] <0.22797.678> closing AMQP connection <0.22797.678> (x.x.x.x05:57676 -> x.x.x.x00:5672 - neutron-openvswitch-agent:7:78d30277-173a-47d4-b8a8-3edde75f115e): missed heartbeats from client, timeout: 60s 2021-09-17 06:43:09.817 [info] <0.7903.679> Closing all channels from connection 'x.x.x.x05:57676 -> x.x.x.x00:5672' because it has been closed 2021-09-17 06:43:10.319 [error] <0.18268.678> closing AMQP connection <0.18268.678> (x.x.x.x00:44924 -> x.x.x.x00:5672 - nova-conductor:957:e3951cbd-37ca-4d9f-8fac-394c83b36f09): missed heartbeats from client, timeout: 60s 2021-09-17 06:43:10.322 [info] <0.28032.678> Closing all channels from connection 'x.x.x.x00:44924 -> x.x.x.x00:5672' because it has been closed 2021-09-17 06:43:10.330 [info] <0.19596.678> accepting AMQP connection <0.19596.678> (x.x.x.x00:46850 -> x.x.x.x00:5672) 2021-09-17 06:43:10.333 [info] <0.19596.678> Connection <0.19596.678> (x.x.x.x00:46850 -> x.x.x.x00:5672) has a client-provided name: nova-conductor:957:e3951cbd-37ca-4d9f-8fac-394c83b36f09 2021-09-17 06:43:10.334 [info] <0.19596.678> connection <0.19596.678> (x.x.x.x00:46850 -> x.x.x.x00:5672 - nova-conductor:957:e3951cbd-37ca-4d9f-8fac-394c83b36f09): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:10.831 [info] <0.9838.678> accepting AMQP connection <0.9838.678> (x.x.x.x05:57878 -> x.x.x.x00:5672) 2021-09-17 06:43:10.834 [info] <0.9838.678> Connection <0.9838.678> (x.x.x.x05:57878 -> x.x.x.x00:5672) has a client-provided name: neutron-openvswitch-agent:7:78d30277-173a-47d4-b8a8-3edde75f115e 2021-09-17 06:43:10.835 [info] <0.9838.678> connection <0.9838.678> (x.x.x.x05:57878 -> x.x.x.x00:5672 - neutron-openvswitch-agent:7:78d30277-173a-47d4-b8a8-3edde75f115e): user 'openstack' authenticated and granted access to vhost '/' 2021-09-17 06:43:11.738 [error] <0.15275.678> closing AMQP connection <0.15275.678> (x.x.x.x01:53414 -> x.x.x.x00:5672 - mod_wsgi:968:b60f723c-882e-4780-a162-55c03ebe6179): missed heartbeats from client, timeout: 60s /var/log/kolla/glance/glance-api.log 2021-09-17 13:17:23.212 958 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:17:29.208 955 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:17:42.190 957 INFO eventlet.wsgi.server [-] x.x.x.x00 - - [17/Sep/2021 13:17:42] "GET / HTTP/1.1" 300 1446 0.001949 2021-09-17 13:17:46.351 957 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:18:12.336 955 INFO eventlet.wsgi.server [-] x.x.x.x00 - - [17/Sep/2021 13:18:12] "GET / HTTP/1.1" 300 1446 0.003754 2021-09-17 13:18:13.703 956 INFO eventlet.wsgi.server [req-1a00db34-349b-4ed3-8625-edfadc939be4 d33b56f82a984f6a99bad58717f6f0ef 6ee7e3517d254f9e9ee5eac7d56a3179 - default default] x.x.x.x09,x.x.x.x04 - - [17/Sep/2021 13:18:13] "GET /v2/images?limit=20 HTTP/1.1" 200 2990 0.139229 2021-09-17 13:18:33.065 955 INFO eventlet.wsgi.server [req-a57046eb-549e-4837-95c7-d0994c05449e d33b56f82a984f6a99bad58717f6f0ef 6ee7e3517d254f9e9ee5eac7d56a3179 - default default] x.x.x.x09,x.x.x.x04 - - [17/Sep/2021 13:18:33] "GET /v2/images?limit=20 HTTP/1.1" 200 2990 0.096302 2021-09-17 13:18:33.077 955 INFO eventlet.wsgi.server [req-3a8af281-6732-4afd-b3ec-e502367bce51 d33b56f82a984f6a99bad58717f6f0ef 6ee7e3517d254f9e9ee5eac7d56a3179 - default default] x.x.x.x09,x.x.x.x04 - - [17/Sep/2021 13:18:33] "GET /v2/schemas/image HTTP/1.1" 200 6259 0.003982 2021-09-17 13:18:36.747 959 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:18:36.759 959 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:18:42.460 957 INFO eventlet.wsgi.server [-] x.x.x.x00 - - [17/Sep/2021 13:18:42] "GET / HTTP/1.1" 300 1446 0.001383 2021-09-17 13:18:52.659 956 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:18:52.670 956 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:19:12.598 955 INFO eventlet.wsgi.server [-] x.x.x.x00 - - [17/Sep/2021 13:19:12] "GET / HTTP/1.1" 300 1446 0.001590 2021-09-17 13:19:16.367 957 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:19:27.127 955 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:19:27.184 955 INFO eventlet.wsgi.server [req-8f68efc9-f46a-450d-8138-928e4f2585b4 d33b56f82a984f6a99bad58717f6f0ef 6ee7e3517d254f9e9ee5eac7d56a3179 - default default] x.x.x.x09,x.x.x.x04 - - [17/Sep/2021 13:19:27] "GET /v2/images?limit=20 HTTP/1.1" 200 2990 0.032917 /var/log/kolla/keystone/keystone.log 2021-09-17 13:13:54.670 958 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:13:55.148 960 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:13:55.150 962 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:13:55.345 959 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:13:56.044 961 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:14:18.402 966 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:14:31.372 963 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:14:36.580 967 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:14:38.771 965 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:16:12.226 964 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer /var/log/kolla/nova/nova-compute.log 2021-09-17 13:10:07.710 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 56226. 2021-09-17 13:11:11.473 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:11:34.046 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:11:34.721 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:13:07.699 7 ERROR oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] AMQP server on x.x.x.x00:5672 is unreachable: Server unexpectedly closed connection. Trying again in 1 seconds.: OSError: Server unexpectedly closed connection 2021-09-17 13:13:08.734 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 56434. 2021-09-17 13:14:11.501 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:14:34.070 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:14:34.745 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:16:08.724 7 ERROR oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] AMQP server on x.x.x.x00:5672 is unreachable: Server unexpectedly closed connection. Trying again in 1 seconds.: OSError: Server unexpectedly closed connection 2021-09-17 13:16:09.756 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 56638. 2021-09-17 13:17:34.098 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:19:09.747 7 ERROR oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] AMQP server on x.x.x.x00:5672 is unreachable: Server unexpectedly closed connection. Trying again in 1 seconds.: OSError: Server unexpectedly closed connection 2021-09-17 13:19:10.781 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 56834. 2021-09-17 13:20:34.124 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:20:34.777 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:20:51.789 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:22:10.772 7 ERROR oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] AMQP server on x.x.x.x00:5672 is unreachable: Server unexpectedly closed connection. Trying again in 1 seconds.: OSError: Server unexpectedly closed connection 2021-09-17 13:22:11.806 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 57038. 2021-09-17 13:23:34.149 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:23:34.803 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:23:51.818 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:25:11.795 7 ERROR oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] AMQP server on x.x.x.x00:5672 is unreachable: Server unexpectedly closed connection. Trying again in 1 seconds.: OSError: Server unexpectedly closed connection 2021-09-17 13:25:12.842 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 57240. 2021-09-17 13:26:34.176 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:26:34.830 7 INFO oslo.messaging._drivers.impl_rabbit [-] A recoverable connection/channel error occurred, trying to reconnect: [Errno 104] Connection reset by peer 2021-09-17 13:28:12.812 7 ERROR oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] AMQP server on x.x.x.x00:5672 is unreachable: Server unexpectedly closed connection. Trying again in 1 seconds.: OSError: Server unexpectedly closed connection 2021-09-17 13:28:13.848 7 INFO oslo.messaging._drivers.impl_rabbit [-] [9d51e0ec-bd43-46b4-8c57-a9c6aa01903d] Reconnected to AMQP server on x.x.x.x00:5672 via [amqp] client with port 57446.
-- Hervé Beraud Senior Software Engineer at Red Hat irc: hberaud https://github.com/4383/ https://github.com/4383/ https://twitter.com/4383hberaud https://twitter.com/4383hberaud
On 9/21/21 2:44 AM, bkslash wrote:
So there are two questions:
- Why heartbeat_in_pthread didn’t change situation? And if it’s
deprecated it was kind of risky to use it long term (after upgrade to wallaby it would stop working anyway?)
It's deprecated because it became the default behavior in subsequent releases. The functionality is not going away, just the ability to turn it off.
The only reason it was ever configurable was that we had concerns about mixing pthreads and greenthreads and wanted the people affected by the original bug to test the combination for a couple of releases before we turned it on for everyone.
On Wed, 2021-09-22 at 11:22 -0500, Ben Nemec wrote:
On 9/21/21 2:44 AM, bkslash wrote:
So there are two questions:
- Why heartbeat_in_pthread didn’t change situation? And if it’s
deprecated it was kind of risky to use it long term (after upgrade to wallaby it would stop working anyway?)
It's deprecated because it became the default behavior in subsequent releases. The functionality is not going away, just the ability to turn it off.
The only reason it was ever configurable was that we had concerns about mixing pthreads and greenthreads and wanted the people affected by the original bug to test the combination for a couple of releases before we turned it on for everyone.
actully the ablity to trun it off is not going away either turing it on can break the nova comptue agent by breaking eventlet. https://bugs.launchpad.net/oslo.messaging/+bug/1934937 here is the patch to undeprecate it https://review.opendev.org/c/openstack/oslo.messaging/+/800621 based on the bug above this looke like it should be turned off unless you runing under mod_wsgi or uwsgi where the pthread is needed to escape there treadign lifecycle.
On 9/22/21 2:05 PM, Sean Mooney wrote:
On Wed, 2021-09-22 at 11:22 -0500, Ben Nemec wrote:
On 9/21/21 2:44 AM, bkslash wrote:
So there are two questions:
- Why heartbeat_in_pthread didn’t change situation? And if it’s
deprecated it was kind of risky to use it long term (after upgrade to wallaby it would stop working anyway?)
It's deprecated because it became the default behavior in subsequent releases. The functionality is not going away, just the ability to turn it off.
The only reason it was ever configurable was that we had concerns about mixing pthreads and greenthreads and wanted the people affected by the original bug to test the combination for a couple of releases before we turned it on for everyone.
actully the ablity to trun it off is not going away either turing it on can break the nova comptue agent by breaking eventlet. https://bugs.launchpad.net/oslo.messaging/+bug/1934937 here is the patch to undeprecate it https://review.opendev.org/c/openstack/oslo.messaging/+/800621 based on the bug above this looke like it should be turned off unless you runing under mod_wsgi or uwsgi where the pthread is needed to escape there treadign lifecycle.
Ah, okay. I thought I remembered seeing some discussion about that but I missed the patch. +2.
Well, then see what I have in impl_rabbit.py (kolla-ansible 11, newest container images, Victoria):
cfg.BoolOpt('heartbeat_in_pthread', default=False, help="EXPERIMENTAL: Run the health check heartbeat thread " "through a native python thread. By default if this " "option isn't provided the health check heartbeat will " "inherit the execution model from the parent process. By " "example if the parent process have monkey patched the " "stdlib by using eventlet/greenlet then the heartbeat " "will be run through a green thread."),
1. It is disabled by default, so heartbeats run through green thread? 2. Why enabling it didn’t change anything? 3. heartbeat = 600 for rabbit resolved problem for now (but I’ve seen 1-2 timeouts after 2 days anyway), but 10-minute timeout is kind of big, so I would like to use another solution
Best regards Adam Tomas
Wiadomość napisana przez Sean Mooney smooney@redhat.com w dniu 22.09.2021, o godz. 21:05:
On Wed, 2021-09-22 at 11:22 -0500, Ben Nemec wrote:
On 9/21/21 2:44 AM, bkslash wrote:
So there are two questions:
- Why heartbeat_in_pthread didn’t change situation? And if it’s
deprecated it was kind of risky to use it long term (after upgrade to wallaby it would stop working anyway?)
It's deprecated because it became the default behavior in subsequent releases. The functionality is not going away, just the ability to turn it off.
The only reason it was ever configurable was that we had concerns about mixing pthreads and greenthreads and wanted the people affected by the original bug to test the combination for a couple of releases before we turned it on for everyone.
actully the ablity to trun it off is not going away either turing it on can break the nova comptue agent by breaking eventlet. https://bugs.launchpad.net/oslo.messaging/+bug/1934937 here is the patch to undeprecate it https://review.opendev.org/c/openstack/oslo.messaging/+/800621 based on the bug above this looke like it should be turned off unless you runing under mod_wsgi or uwsgi where the pthread is needed to escape there treadign lifecycle.
participants (4)
-
Adam Tomas
-
Ben Nemec
-
bkslash
-
Sean Mooney