Closed saingithub closed 2 years ago
In my kubernetes environment Dell CSI driver is the storage provisioner. When I implemented this PV-exporter I am getting the following error in the logs No block storage pvc found or not supported yet
@saingithub Could you please share the result for "df -h" ? and tell me which version you using.
Thanks, @kais271, It worked for me. Could you please let me know how can I add it in Grafana Dashboard
@saingithub which version is you using ? If you using v0.1.2, you can try use this query "(sum without (container,pod,service,namespace,job,instance,endpoint) (pvc_usage)) + on(volumename) group_left(persistentvolumeclaim,mountedby,pod_namespace)pvc_mapping*0 "
The version i am using is v7.1.5. When I give that query I am getting the following error.
"invalid expression type \"string\" for range query, must be Scalar or instant Vector"
The version i am using is v7.1.5. When I give that query I am getting the following error.
"invalid expression type "string" for range query, must be Scalar or instant Vector"
don't have "" ..just *(sum without (container,pod,service,namespace,job,instance,endpoint) (pvc_usage)) + on(volumename) group_left(persistentvolumeclaim,mountedby,pod_namespace)pvc_mapping0**
"many-to-many matching not allowed: matching labels must be unique on one side"
could you please share the full error? and check whether they have the same "job"
Is there is any option to get it in a tabular format in Grafana
Sorry, I don't know
I tried deploying v0.1.3-alpha version. I am getting the following error in the pod log
File "./pvc-exporter.py", line 9, in
you can check if the yaml of the pod has an EXPORTER_SERVER_PORT variable
I had used the yaml file of v0.1.2 for the deployment of the pod.
that's not work. you need use v0.1.3-alpha chart to install v0.1.3-alpha. because it has more changes.
In v0.1.3 you have now created deployment files in which lots of variables are used compared to the previous version. How can I interpret that to deploy in a Kubernetes environment
@saingithub There are 3 variables: HOST_IP,EXPORTER_SERVER_PORT,SCAN_INTERVAL. HOST_IP: This used to match pv. We get all pv via API, and deploy pvc-exporter as daemonset. So the pod only get the pv size of hostpath pv on current the node. EXPORTER_SERVER_PORT: This is exporter http server port in pod. SCAN_INTERVAL: This is scan pvc interval, unit is second.
I have more than 100 PVC. I have attached the grafana dashboard image. Is there any way to filter it more so that we can interpret it easy.
@saingithub You can try this dashsboard. It can filter by namespace or pod. https://github.com/kais271/pvc-exporter/blob/v0.1.3-alpha/docs/PVC%20%20Dashboard-1637745492005.json
When I used the above file I am getting the following error in Grafana
Panel plugin not found: timeseries
Maybe caused by grafana version. My version is v8.1.2 . Try the following steps
(sum without (container,pod,service,namespace,job,instance,endpoint,pvc_namespace,pvc_requested_size_MB) (pvc_usage)) + on(persistentvolume) group_left(persistentvolumeclaim,mountedby,pod_namespace)pvc_mapping{pod_namespace=~"$pod_namespace",mountedby=~"$Pod"}*0
2.
3.
When i give the above expression no data is getting in the dashboard
Hi, Thanks, @kais271.
It worked for me. Now I was able to differentiate PVC usage by namespace and pods.
One more thing , is there any way to create an alert mechanism if any PVC exceeds a specified limit
@saingithub You can reference this: https://grafana.com/docs/grafana/latest/alerting/old-alerting/create-alerts/ But "Template variables are not supported in alert queries", so you need to create a new dashboard for alert.
Issues close after 30d of inactivity. Reopen the issue with /reopen.
In my kubernetes environment Dell CSI driver is the storage provisioner. When I implemented this PV-exporter I am getting the following error in the logs No block storage pvc found or not supported yet