Jira (PDB-4444) PuppetDB never finishes migrating resource_events

Luke Bigum (JIRA) Wed, 14 Aug 2019 01:51:20 -0700

Title: Message Title

Luke Bigum commented on

For the JVM, we only change the Max Heap:

JAVA_ARGS="-Xmx6g"

The current row count (which is a higher than I was expecting):

puppetdb=> select count(*) from resource_events;

 count

----------

42973699

(1 row)

The machine itself is a 12 core KVM instance with 40Gig total RAM. The Postgresql instance is co-located on the same VM, and it also runs a Puppet Server instance (but we don't have automatic Puppet runs enabled so this Puppet Server is mostly idle).

Some other relevant config:

# How often (in minutes) to compact the database

# gc-interval = 60

gc-interval = 60

Number of seconds before any SQL query is considered 'slow'; offending

# queries will not be interrupted, but will be logged at the WARN log level.

log-slow-statements = 10

syntax_pgs = true

node-ttl = 0s

node-purge-ttl = 1d

report-ttl = 14d

conn-max-age = 60

conn-keep-alive = 45

conn-lifetime = 0

We don't purge any nodes based on last check-in time, but I do try purge reports. We can't turn on node-ttl because our Puppet runs are done on a certain schedule. If we were to turn on automatic purging, we might expire an active node simply because it hasn't had it's scheduled run. This affects our monitoring, which is derived from PuppetDB. When a machine is decommissioned it is deactivated in PuppetDB manually.

I thought purging reports would be enough to keep `resource_events` low, but could be it's not doing what I think it's doing.

Add Comment

This message was sent by Atlassian JIRA

--
You received this message because you are subscribed to the Google Groups "Puppet Bugs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-bugs/JIRA.314156.1561543112000.55104.1565772660358%40Atlassian.JIRA.

Jira (PDB-4444) PuppetDB never finishes migrating resource_events

Reply via email to