openstack
/
fuel-astute
Code Issues Proposed changes

Don't use retries for puppet deployments by default 

Granular deployment task with possibility for user
to control number of retries for any task now allow
Astute to decrease number of retries from 1 to 0 by default.

It will give us such benefits:

* fail fast: do not use useless retries for failed task;
* detect cases where we need to place retries inside task code;
* show us tasks with bad design from box.

For special it could be setup by hands.

Change-Id: I0b5645bdceb524f3b0e0135125c9a3f457fbc3b4
Closes-Bug: #1457887
This commit is contained in:
Vladimir Sharshov (warpc) 2015-09-22 16:11:46 +03:00
parent 0f753467a3
commit c145db3476
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
lib/astute/config.rb
View File

@ -58,7 +58,7 @@ module Astute
conf[:puppet_timeout] = 90 * 60 # maximum time it waits for single puppet run
conf[:puppet_deploy_interval] = 2 # sleep for ## sec, then check puppet status again
conf[:puppet_fade_timeout] = 120 # how long it can take for puppet to exit after dumping to last_run_summary
conf[:puppet_retries] = 2 # how many times astute will try to run puppet
conf[:puppet_retries] = 0 # how many times astute will try to run puppet
conf[:mc_retries] = 10 # MClient tries to call mcagent before failure
conf[:mc_retry_interval] = 1 # MClient sleeps for ## sec between retries
conf[:puppet_fade_interval] = 30 # retry every ## seconds to check puppet state if it was running