yahoo-eng-team team mailing list archive

[Simon <1263745@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

The "timeout check 5s" option is causing a 5 second delay with a chance
of 1/N when one of the N backend nodes is down.

backend 36279594-bd3f-47dd-b2fa-96e0bfc02ac6
        mode tcp
        balance roundrobin
        timeout check 5s
        server 7ad07521-3034-4e1c-9b3c-33ac42ac50d8 172.20.144.74:80 weight 1 check inter 3s fall 5
        server b87715a5-853d-412d-9d74-ed48f83e5fc4 172.20.144.75:80 weight 1 check inter 3s fall 5

Removing this option gives the correct behavior that the downed node is
simply ignored, instead of still being redirected to and causing 5s
delay.

** Affects: neutron
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to neutron.
https://bugs.launchpad.net/bugs/1263745

Title:
  lbaas-haproxy default config causing long delay with downed member

Status in OpenStack Neutron (virtual network service):
  New

Bug description:
  The "timeout check 5s" option is causing a 5 second delay with a
  chance of 1/N when one of the N backend nodes is down.

  backend 36279594-bd3f-47dd-b2fa-96e0bfc02ac6
          mode tcp
          balance roundrobin
          timeout check 5s
          server 7ad07521-3034-4e1c-9b3c-33ac42ac50d8 172.20.144.74:80 weight 1 check inter 3s fall 5
          server b87715a5-853d-412d-9d74-ed48f83e5fc4 172.20.144.75:80 weight 1 check inter 3s fall 5

  Removing this option gives the correct behavior that the downed node
  is simply ignored, instead of still being redirected to and causing 5s
  delay.

To manage notifications about this bug go to:
https://bugs.launchpad.net/neutron/+bug/1263745/+subscriptions