Open Stack

Please keep the responses in the list so others can see your  
responses, too, and chime in with their comments.

How is your dashboard configured regarding cache? This is our config  
(/srv/www/openstack-dashboard/openstack_dashboard/local/local_settings.d/_100_local_settings.py):

CACHES = {
     'default': {
         'BACKEND': 'django.core.cache.backends.memcached.MemcachedCache',
         'LOCATION': [
           '<IP1>:11211',
           '<IP2>:11211',
         ]
     }
}

Our memcached config has both localhost and its management IP:

MEMCACHED_PARAMS="-U 0 -m 64 -l 127.0.0.1,<IP1> -p 11211 -c 4096"

The reason for that is we noticed performance improvements when we  
replaced the list of control nodes for all the services and point to  
localhost instead:

# before
#memcached_servers = controller01:11211,controller02:11211

# after
memcached_servers = localhost:11211



Zitat von Adivya Singh <adivya1.singh at gmail.com>:

> Hello,
>
> My setting in memcache looks like this, because it is default taken by
> ansible
>
> -d
> -v
> -U 0
> -m 1024
> -p 11211
> -u memcache
> -l 172.29.x.X
> -c 4096
> -t 4
>
> So should i change the time out perimeter to something else
>
> Regards
> Adivya Singh
>
> On Wed, Jun 15, 2022 at 8:53 PM Maysa De Macedo Souza <mdemaced at redhat.com>
> wrote:
>
>> Hello,
>>
>> Maybe increasing the value of a setting like
>> memcache_pool_conn_get_timeout can help?
>>
>> Cheers,
>> Maysa.
>>
>> On Wed, Jun 15, 2022 at 2:05 PM Eugen Block <eblock at nde.ag> wrote:
>>
>>> Hi,
>>>
>>> I remember having a similar issue a few years back in an older
>>> openstack release, probably Ocata. We had difficulties loading *all*
>>> instances in the admin tab in Horizon dashboard. It turned out to be a
>>> memcached misconfiguration. I don't recall the details, unfortunately.
>>> But I'd start with memcached.
>>>
>>>
>>> Zitat von Adivya Singh <adivya1.singh at gmail.com>:
>>>
>>> > hi Team,
>>> >
>>> > We have Xena release installed on Ubuntu Openstack, Lately we are
>>> getting
>>> > "504 Gateway Timeout error ", As the we moved from 1 node to 3 node Open
>>> > Stack
>>> > i do not see any error in haproxy , neither there is a time out when i
>>> ping
>>> > external-VIP-IP configured.
>>> >
>>> > Apache 2 are also loaded, tried to delete the Horizon Container and
>>> created
>>> > it again, but no errors, Still the same errors.
>>> >
>>> > Any guesses on that
>>> >
>>> > Regards
>>> > Adivya Singh
>>>
>>>
>>>
>>>
>>>