Consider increasing the configuration parameter "max_wal_size"

Product Support

AI-generated summary

@diegoeis encountered an error while trying to rebuild their Discourse app after changing the SMTP address in app.yml. The error message suggested increasing the configuration parameter “max_wal_size” due to frequent checkpoints.

@diegoeis claims to have solved the issue by following the commands provided by @codinghorror in a related post.

@codinghorror responded, asking if the issue was caused by running out of disk space, as the wal errors are related to Postgres writing its database change log to disk.

1

Hi, I’m trying change the smtp address of my smtp email server. I edit my app.yml, change the address, :wq to save, and try rebuild the app. In a specific point, the system warn me this error.

Using spork-rails 4.0.0
0:  spork-rails (4.0.0) from /var/www/discourse/vendor/bundle/ruby/2.3.0/specifications/spork-rails-4.0.0.gemspec
Bundle complete! 99 Gemfile dependencies, 179 gems now installed.
Gems in the group development were not installed.
Bundled gems are installed into ./vendor/bundle.

I, [2016-07-10T04:04:40.895077 #13]  INFO -- : > cd /var/www/discourse && su discourse -c 'bundle exec rake db:migrate'
2016-07-10 04:05:40 UTC [46-1] LOG:  checkpoints are occurring too frequently (24 seconds apart)
2016-07-10 04:05:40 UTC [46-2] HINT:  Consider increasing the configuration parameter "max_wal_size".
2016-07-10 04:06:00 UTC [46-3] LOG:  checkpoints are occurring too frequently (20 seconds apart)
2016-07-10 04:06:00 UTC [46-4] HINT:  Consider increasing the configuration parameter "max_wal_size".
2016-07-10 04:06:19 UTC [46-5] LOG:  checkpoints are occurring too frequently (19 seconds apart)
2016-07-10 04:06:19 UTC [46-6] HINT:  Consider increasing the configuration parameter "max_wal_size".
2016-07-10 04:06:36 UTC [46-7] LOG:  checkpoints are occurring too frequently (17 seconds apart)
2016-07-10 04:06:36 UTC [46-8] HINT:  Consider increasing the configuration parameter "max_wal_size".
2016-07-10 04:06:53 UTC [46-9] LOG:  checkpoints are occurring too frequently (17 seconds apart)
2016-07-10 04:06:53 UTC [46-10] HINT:  Consider increasing the configuration parameter "max_wal_size".
2

I think I solved this issue making the commands given by @codinghorror in this post:
https://meta.discourse.org/t/discourse-using-almost-95-of-40gb-total-drive-space-is-this-a-problem/43966/2?u=diegoeis

3

Were you out of disk space? The wal stuff has to do with postgres streaming its real time log of database changes to disk.