CLI Commands
Maximo-CPI offers a set of CLI commands to collect metrics across various categories, including node-level, cluster-level, certificate-related data, pod-level, application-level, performance alerts. The command set can be used for the real time diagnosis, alert/notification and more.
Command output
Most commands support two output formats: standard screen column print (default) and JSON format.
JSON format requires an argument json (e.g., get-node-cpu.sh json).
Acceptable Env Variables
Note: Environment Variables are Case-Sensitive. Use unset DTR to remove the variables
| Env Variable Name | Description | Default Value |
|---|---|---|
| INTERVALLIST | Default intervallist | export INTERVALLIST="5m 15m 1h 3h 1d 3d 5d" |
| NSLIST | Namespace list related to MAS apps | export NSLIST="mas-.|db2u|.mongo.|.-sls|kmodels.|aibroker." |
| EXPIREDINDAYS | Check the certficate expiration in how many days | export EXPIREDINDAYS=10 |
| DTR | Default Time Range note: If the command accepts both DTR and INTERVALLIST, it will ignore INTERVALLIST when DTR is set. |
No default value. Sample command: export DTR="1h" |
| TP_DTR | Default Time Range for get_toppod.sh | export TP_DTR="1h" |
| TP_TOPK | The maximum number of records returned for the top pods. | export TP_TOPK="20" |
Metric-Related Command List
| CLI Name | Description | Acceptable Env Variable |
|---|---|---|
| get-all-pod-info.sh | get the pod info e.g. age, status, restart, cpu request/limit, mem request/limit from all namespaces. | None |
| get-cert-info.sh | get the expired certificate | EXPIREDINDAYS |
| get-cluster-info.sh | get the info for clusterversion, node count, podcount, cpuAlloc,cpuUtil ,cpuUtil_max_5d ,cpuRC, cpuLC,memAlloc,memUtil ,memUtil_max_5d ,memRC, memLC | None |
| get-ephemeral-storage.sh | get the ephemeral storage info e.g. instance, device,total_gb,avail_gb,used_gb util_percentage | None |
| get-ingressor-info.sh | get the maxconn value of the ingressor controller | None |
| get-mas-info.sh | get mas deployment info | None |
| get-node-cpu.sh | get the node cpu metrcis | INTERVALLIST |
| get-node-info.sh | get the node spec | INTERVALLIST |
| get-node-mem.sh | get the node memory metrics | INTERVALLIST |
| get-node-net.sh | get the node network metrics | INTERVALLIST |
| get-node-pidslimit.sh | get the node pid limits per container | None |
| get-ns-info.sh | get the namespace cpu_request mem_request mem_limits cpu_limits for all namespaces | None |
| get-ns-metric.sh | get the resource metrics for mas related namespaces | DTR (or INTERVALLIST), NSLIST |
| get-pod-cpu.sh | get the pod cpu metrics for mas related namespaces | DTR (or INTERVALLIST), NSLIST |
| get-pod-fs.sh | get the pod filesystem metrics for mas related namespaces | DTR (or INTERVALLIST), NSLIST |
| get-pod-info.sh | get the pod spec info | None |
| get-pod-mem.sh | get the pod mem metrics for mas related namespaces | DTR (or INTERVALLIST), NSLIST |
| get-pod-net.sh | get the pod network metrics for mas related namespaces | DTR (or INTERVALLIST), NSLIST |
| get-pod-restart.sh | get the pod restart metrics for mas related namespaces | DTR (or INTERVALLIST), NSLIST |
| get-pvc-info.sh | get all PVC info | None |
| get-toppods.sh | get the top pods fron | TP_DTR and TP_TOPK |
Filter the json output by JQ query
Consider to use the jq command to filter the output from the CLI. Here’s an example:
- list the nodes whose max_cpu_1h > 40:
get-node-cpu.sh json|jq '[.[] | select(.max_cpu_1h > 40)]'