Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Performance data writer causes runaway-series-cardinality in InfluxDB #531

Open
slalomsk8er opened this issue Sep 20, 2023 · 0 comments
Open

Comments

@slalomsk8er
Copy link
Contributor

Expected Behavior

Performance data writer is "setup and forget" and provides a clear benefit.

Current Behavior

  • In large and highly dynamic vCenters (virtual desktop infrastructure) the writer can result in runaway-series-cardinality:
Sep 15 12:02:11 icingalp06 influxd-systemd-start.sh[6778]: ts=2023-09-15T10:02:11.985985Z lvl=warn msg="max-values-per-tag limit may be exceeded soon" log_id=0kFnS2vG000 service=store perc=100% n=100000 max=100000 db_instance=icinga_vspheredb_ictvdivclp01 measurement=VmMemory tag=vm_uuid

[12:03:35 root@icingalp06 ~]$ journalctl -xeu influxdb | grep max-values-per-tag | cut -d' ' -f 19 | sort -u
db_instance=icinga_vspheredb_ictvdivclp01

[12:04:08 root@icingalp06 ~]$ journalctl -xeu influxdb | grep max-values-per-tag | cut -d' ' -f 20 | sort -u
measurement=VirtualDisk
measurement=VirtualNetworkAdapter
measurement=VmCpu
measurement=VmMemory

[12:04:15 root@icingalp06 ~]$ journalctl -xeu influxdb | grep max-values-per-tag | cut -d' ' -f 21 | sort -u
tag=instance
tag=vm_moref
tag=vm_uuid

https://www.influxdata.com/blog/solving-runaway-series-cardinality-when-using-influxdb/

  • Also in the module itself, there isn't a clear benefit for setting all this up. Including "fixing" the Grafana dashboards.

So for the time being, we disabled the feature and droppend the InfluxDB icinga_vspheredb_* instances.
Because this feature endangers the Grafana graphs in the monitoring module and I suspect nobody except me ever opened the Grafana dashboard belonging to this module.

Possible Solution

  • Performance data writer toggle per vCenter
  • Performance data writer sends data in a way to not trigger runaway-series-cardinality
  • Add Grafana graphs to the module and not just display the current state of CPU, RAM and the like.

Your Environment

  • VMware vCenter®/ESXi™-Version: vCenter Server 6.7.0
  • Version/GIT-Hash of this module: 1.6.0
  • Icinga Web 2 version: 2.11.4
  • Operating System and version: AlmaLinux8
  • Webserver, PHP versions: PHP 8.0.30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant