Thanks Zachary that doc is a great help. I'll lower the node-purge-ttl and see how it performs.
On Friday, 4 October 2019 22:14:04 UTC+1, Zachary Kent wrote: > > Hi Andy, > > Are you seeing the deactivated nodes' catalogs turn up in query results > from PDB or only when you query the postgres table directly? > > I'm wondering if you might be hitting some strangeness around the > node-purge-ttl > <https://puppet.com/docs/puppetdb/6.3/configure.html#node-purge-ttl> > setting. This setting specifies the amount of time that the PDB garbage > collection process will wait before permanently deleting a node even if it > has been deactivated or expired. This setting is needed on the enterprise > side of things but could be preventing the nodes that you have deactivated > from being deleted right away. > > I tried setting this value to node-purge-ttl=1s locally and was able to > confirm that deactivating a node and then sending a purge_nodes command > <https://puppet.com/docs/puppetdb/6.3/api/admin/v1/cmd.html#post-pdbadminv1cmd> > > to the admin endpoint triggered gc and deleted the node. > > Note that if you do try this approach you may want to consider using a > batch_limit for the purge_nodes admin command. Otherwise the PDB will try > to delete everything at once which may take a while. See this blog post > <https://puppet.com/blog/preventative-maintenance-for-puppetdb-node-purge-ttl> > > that discusses some of the related issues. > > Changing this setting will also speed up the purging of data from nodes > that have fallen inactive for longer than node-ttl > <https://puppet.com/docs/puppetdb/6.3/configure.html#node-ttl> which is > another thing to consider if you care about querying for nodes that have > stopped checking in longer than node-ttl. > > Hope this helps! > > > On Thu, Oct 3, 2019 at 10:25 AM Andy Hall <[email protected] > <javascript:>> wrote: > >> hey there we have just migrated hundreds of hosts from 3.x to 6.x and >> although lots of work we are almost home and dry but have an issue with >> puppetdb which I hope can be solved. we are running puppetdb-6.3.4 but when >> removing an old node as follows: >> >> puppet node deactivate <nodename> >> >> the information is _not_ getting removed from puppetdb and we have to run >> the following sql manually: >> >> delete from catalogs where certname in (select certname from certnames >> where deactivated is not null); >> >> this is far from ideal and we really need this functionality to work >> again as we have numerous exported resources such as nagios which have to >> be removed when decommissioning a host. >> >> please advise if this is a known issue or if we are doing something wrong. >> >> thanks very much and keep up the good work !! >> >> -- >> You received this message because you are subscribed to the Google Groups >> "Puppet Users" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <javascript:>. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/puppet-users/9f3a3c96-29ec-401d-b3d4-cf6b8270535f%40googlegroups.com >> >> <https://groups.google.com/d/msgid/puppet-users/9f3a3c96-29ec-401d-b3d4-cf6b8270535f%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> > -- You received this message because you are subscribed to the Google Groups "Puppet Users" 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-users/6c54e96c-e09f-4e24-9bd1-b09d7c9d8919%40googlegroups.com.
