Hi,
Is there some kind of script that I can run to know the exact issue about what resource crunch is there ?
Below is the memory and disk utilization of controller:
[heat-admin@overcloud-controller-0 ~]$ free -m
total used free shared buff/cache available
Mem: 128722 25632 92042 70 11047 102377
Swap: 0 0 0
[heat-admin@overcloud-controller-0 ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
devtmpfs 63G 0 63G 0% /dev
tmpfs 63G 39M 63G 1% /dev/shm
tmpfs 63G 27M 63G 1% /run
tmpfs 63G 0 63G 0% /sys/fs/cgroup
/dev/sda2 1.9T 12G 1.9T 1% /
tmpfs 13G 0 13G 0% /run/user/0
tmpfs 13G 0 13G 0% /run/user/1000
Same memory and disk available on all three controllers.
On the same environment when I installed overcloud with redhat repos and redhat overcloud images...I have not faced this issue. I have tested almost 500 projects and 500 networks (one network per project) on the same environment with redhat and it was working fine without any issue and cluster failure. But when I am using Centos-7 tripleo repo's its happening again n again.
Regards
Rahul Pathak
i2k2 Networks (P) Ltd. | Spring Meadows Business Park
A61-B4 & 4A First Floor, Sector 63, Noida - 201 301
ISO/IEC 27001:2005 & ISO 9001:2008 Certified
From: "Alfredo Moralejo Alonso" <amoralej@redhat.com>
To: "Arkady Shtempler" <ashtempl@redhat.com>
Cc: "Rahul Pathak" <rpathak@i2k2.com>, "RDO Developmen List" <dev@lists.rdoproject.org>
Sent: Thursday, July 23, 2020 5:40:42 PM
Subject: Re: [rdo-dev] tripleo cluster failure
Hi all!
Rahul - there is nothing relevant in the attached file, you've probably executed LogTool on "working environment", so there is nothing interesting in it.
I think that you had to mention the Error we've detected in an already "crushed" environment, just as I was suggesting you to do.
Yes, nothing interesting in the attached logs.
Alfredo - this Error was logged almost on each OC node at the same time when the problems had started.
hyp-0
------------------------------ LogPath: /var/log/containers/neutron/openvswitch-agent.log.1 ------------------------------
IsTracebackBlock:False
UniqueCounter:1
AnalyzedBlockLinesSize:18
26712-2020-07-21 20:07:54.604 54410 INFO neutron.plugins.ml2.drivers.openvswitch.agent.ovs_neutron_agent [req-a2173b45-f18b-45f9-90d8-cb8ab1754332 - -...<--LogTool-LINE IS TOO LONG!
26713-2020-07-21 20:07:54.605 54410 INFO neutron.plugins.ml2.drivers.openvswitch.agent.ovs_neutron_agent [req-a2173b45-f18b-45f9-90d8-cb8ab1754332 - -...<--LogTool-LINE IS TOO LONG!
26714:2020-07-21 20:07:56.363 54410 ERROR oslo.messaging._drivers.impl_rabbit [-] [2f214f0b-84d0-49d4-bcf4-477565903585] AMQP server overcloud-control...<--LogTool-LINE IS TOO LONG!
26715:2020-07-21 20:07:56.364 54410 ERROR oslo.messaging._drivers.impl_rabbit [-] [664475bc-3b39-4ce5-a60e-f010b8d5201d] AMQP server overcloud-control...<--LogTool-LINE IS TOO LONG!
26716:2020-07-21 20:07:56.364 54410 ERROR oslo.messaging._drivers.impl_rabbit [-] [68a0ab43-a216-43fe-aa58-860d3dc5e69e] AMQP server overcloud-control...<--LogTool-LINE IS TOO LONG!
...
...
...
LogTool --> THIS BLOCK IS TOO LONG!
LogTool --> POTENTIAL BLOCK'S ISSUES:
26714:2020-07-21 20:07:56.363 54410 ERROR oslo.messaging._drivers.impl_rabbit [-] [2f214f0b-84d0-49d4-bcf4-477565903585] AMQP server overcloud-controller-2.inter...
ection. Check login credentials: Socket closed: IOError: Socket closed
no 111] ECONNREFUSED. Trying again in 1 seconds.: error: [Errno 111] ECONNREFUSED
rnalapi.i2k2cloud02.com:5672
is unreachable: <AMQPError: unknown error>. Trying again in 1
seconds.: RecoverableConnectionError: <AMQPError: unknown error>
I'd suggest to check rabbitmq and mariadb logs.
AFAIK, there is not a configuration that may limit the number of networks or project, but it may be hitting some resources scarcity that affect the running services. What's the memory sizing and usage of the controllers?
You can find more Error Blocks in the attached file.
Thanks!