Open taleintervenor opened 2 years ago
Hi @taleintervenor!
What is your version of PuppetDB?
Because I don't see this issue in one of my instances with PuppetDB 5.2.8 - the number of resources shown in the Puppetboard roughly matches the number of entries in /opt/puppetlabs/puppet/cache/state/resources.txt.
Hi @taleintervenor!
What is your version of PuppetDB?
Because I don't see this issue in one of my instances with PuppetDB 5.2.8 - the number of resources shown in the Puppetboard roughly matches the number of entries in /opt/puppetlabs/puppet/cache/state/resources.txt.
Our puppetserver manages 2 cluster with different slurm version.
The node mentioned above using slurm 21.08.4, and their slurmdbd node using slurm-slurmdbd-21.08.4-1.el8.x86_64 package. All the compute nodes in this cluster are shown with 34505 resources.total while resources.txt containing 479 lines.
And another cluster using slurm 19.05.7, its compute nodes have 520 lines resources.txt file, while puppetboard shows resources.total is 718.
Sorry, but I don't get what slurm version has to do with what we are talking about here. I was asking about PuppetDB version...
That information can be useful because other users of the same version of PuppetDB as you might chip in if they are seeing this issue too or not. Maybe it's a bug in PuppetDB or PyPuppetDB with this specific PuppetDB version?
Sorry, but I don't get what slurm version has to do with what we are talking about here. I was asking about PuppetDB version...
That information can be useful because other users of the same version of PuppetDB as you might chip in if they are seeing this issue too or not. Maybe it's a bug in PuppetDB or PyPuppetDB with this specific PuppetDB version?
Oh I read it wrong, puppetdb version is puppetdb-7.5.2, with puppetserver-7.3.0 and puppet-agent-7.6.1. Because the problem cluster is newly deployed, its main difference from old cluster is system version (centos 8 vs centos 7) and slurm version, so I wonder if this problem is due to these two differences.
Puppetboard 3.3.0 shows our node has 34k resources as below, which is incredibly large. I wonder how puppetboard count this resources total number? The facts table at the same page shows only 302 entries in total, and /opt/puppetlabs/puppet/cache/state/resources.txt on that node has 479 lines content. Both are far smaller than the 34k number.