Skip to content

Releases: ClusterCockpit/cc-metric-collector

cc-metric-collector-v0.4

15 Mar 15:46
Compare
Choose a tag to compare
Merge branch 'develop' into main

cc-metric-collector-v0.3

03 Mar 12:48
Compare
Choose a tag to compare
Set proper user for files

cc-metric-collector-v0.2

25 Feb 13:55
d98076c
Compare
Choose a tag to compare
Merge current development version into main (#48)

* DiskstatCollector: cast part_max_used metric to int

* Add uint types to GangliaSink and LibgangliaSink

* Use new sink instances to allow multiple of same sink type

* Update sink README and SampleSink

* Use new receiver instances to allow multiple of same receiver type

* Fix metric scope in likwid configuration script

* Mention likwid config script in LikwidCollector README

* Refactor: Embed Init() into New() function

* Refactor: Embed Init() into New() function

* Fix: MetricReceiver uses uninitialized values, when initialization fails

* Use Ganglia configuration (#44)

* Copy all metric configurations from original Ganglia code

* Use metric configurations from Ganglia for some metrics

* Format value string also for known metrics

* Numa-aware memstat collector (#45)

* Add samples for collectors, sinks and receivers

* Ping InfluxDB server after connecting to recognize faulty connections

* Add sink for Prometheus monitoring system (#46)

* Add sink for Prometheus monitoring system

* Add prometheus sink to README

* Add scraper for Prometheus clients (#47)

Co-authored-by: Holger Obermaier <[email protected]>
Co-authored-by: Holger Obermaier <[email protected]>

cc-metric-collector-v0.1

21 Feb 18:06
Compare
Choose a tag to compare

First version of the ClusterCockpit metric collector, a node agent to read data from different sources, process them and send them to a set of sinks.

The collector currently contains collectors for:

  • CPU (OS statistics and hardware performance event measurements)
  • Memory
  • Disk
  • Filesystems like NFS, Lustre or GPFS
  • Network (Ethernet and InfiniBand)
  • Nvidia GPUs

Metrics can be sent to one or more sinks:

If you want to create a tree of metric collectors, you can receive data through an NATS based receiver and forward them to the sinks.