← Back to team overview

yahoo-eng-team team mailing list archive

[Bug 1590042] [NEW] wsgi and eventlet logs flooding on large deployments

 

Public bug reported:

Based on observation in one of our larger lab deployments, keystone
/keystone-all.log is seeing on average multiple logs per seconds. These
are logged at INFO verbosity which is the default logging level in
keystone.conf

Most frequently seen logs:
<DATE> <TIME> <NUMBER> INFO keystone.common.wsgi [-] GET /auth/tokens? 

<DATE> <TIME> <NUMBER> INFO eventlet.wsgi.server [-] <NUMBER> - -
[<DATE> <TIME>] "GET /v <NUMBER> /auth/tokens HTTP/ <NUMBER> " <NUMBER>
. <NUMBER>

Other high runner logs: 
<DATE> <TIME> <NUMBER> INFO keystone.common.wsgi [-] POST /tokens?POST /tokens 

<DATE> <TIME> <NUMBER> INFO eventlet.wsgi.server [-] <NUMBER> - -
[<DATE> <TIME>] "POST /v <NUMBER> /tokens HTTP/ <NUMBER> " <NUMBER> .
<NUMBER>

<DATE> <TIME> <NUMBER> INFO keystone.common.wsgi [-] GET /? 
<DATE> <TIME> <NUMBER> INFO eventlet.wsgi.server [-] <NUMBER> - - [<DATE> <TIME>] "GET / HTTP/ <NUMBER> " <NUMBER> <NUMBER>


These log entries cause frequent churn in keystone-all.log.

We have a small fix in place to move GET/POST logs from
keystone.common.wsgi and eventlet.wsgi.server to debug logs and to
enable these logs when the verbosity is set to debug in keystone.conf

** Affects: keystone
     Importance: Undecided
     Assignee: Kam Nasim (knasim-wrs)
         Status: New

** Changed in: keystone
     Assignee: (unassigned) => Kam Nasim (knasim-wrs)

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Identity (keystone).
https://bugs.launchpad.net/bugs/1590042

Title:
  wsgi and eventlet logs flooding on large deployments

Status in OpenStack Identity (keystone):
  New

Bug description:
  Based on observation in one of our larger lab deployments, keystone
  /keystone-all.log is seeing on average multiple logs per seconds.
  These are logged at INFO verbosity which is the default logging level
  in keystone.conf

  Most frequently seen logs:
  <DATE> <TIME> <NUMBER> INFO keystone.common.wsgi [-] GET /auth/tokens? 

  <DATE> <TIME> <NUMBER> INFO eventlet.wsgi.server [-] <NUMBER> - -
  [<DATE> <TIME>] "GET /v <NUMBER> /auth/tokens HTTP/ <NUMBER> "
  <NUMBER> . <NUMBER>

  Other high runner logs: 
  <DATE> <TIME> <NUMBER> INFO keystone.common.wsgi [-] POST /tokens?POST /tokens 

  <DATE> <TIME> <NUMBER> INFO eventlet.wsgi.server [-] <NUMBER> - -
  [<DATE> <TIME>] "POST /v <NUMBER> /tokens HTTP/ <NUMBER> " <NUMBER> .
  <NUMBER>

  <DATE> <TIME> <NUMBER> INFO keystone.common.wsgi [-] GET /? 
  <DATE> <TIME> <NUMBER> INFO eventlet.wsgi.server [-] <NUMBER> - - [<DATE> <TIME>] "GET / HTTP/ <NUMBER> " <NUMBER> <NUMBER>

  
  These log entries cause frequent churn in keystone-all.log.

  We have a small fix in place to move GET/POST logs from
  keystone.common.wsgi and eventlet.wsgi.server to debug logs and to
  enable these logs when the verbosity is set to debug in keystone.conf

To manage notifications about this bug go to:
https://bugs.launchpad.net/keystone/+bug/1590042/+subscriptions


Follow ups