I am looking for the CPU, Memory and Disk consumption for each Tenant in Openstack,and their relationship by users, instances, flavors in use. Horizon only shows utilization of memory, cpu of a global way. Is it possible to get it with Openstack commands?
My openstack is based on Rocky.
Any ideas will be really appreciated
The only thing I know is
openstack limits show --absolute --project <Project_ID/Tenant_ID>
see also https://docs.openstack.org/python-openstackclient/pike/cli/command-objects/limits.html
In the output you have information like for example totalCoresUsed, which represents the number of cores, which are used by the selected project.
Example:
root@openstack-controller:~# openstack limits show --absolute --project 416f937f505f4ff6b623c48a61228a86
+--------------------------+-------+
| Name | Value |
+--------------------------+-------+
| maxTotalInstances | 10 |
| maxTotalCores | 20 |
| maxTotalRAMSize | 51200 |
| maxSecurityGroups | 10 |
| maxTotalFloatingIps | 10 |
| maxServerMeta | 128 |
| maxImageMeta | 128 |
| maxPersonality | 5 |
| maxPersonalitySize | 10240 |
| maxSecurityGroupRules | 20 |
| maxTotalKeypairs | 100 |
| maxServerGroups | 10 |
| maxServerGroupMembers | 10 |
| totalRAMUsed | 2560 |
| totalCoresUsed | 7 |
| totalInstancesUsed | 7 |
| totalFloatingIpsUsed | 0 |
| totalSecurityGroupsUsed | 1 |
| totalServerGroupsUsed | 0 |
| maxTotalVolumes | 10 |
| maxTotalSnapshots | 10 |
| maxTotalVolumeGigabytes | 1000 |
| maxTotalBackups | 10 |
| maxTotalBackupGigabytes | 1000 |
| totalVolumesUsed | 5 |
| totalGigabytesUsed | 7 |
| totalSnapshotsUsed | 0 |
| totalBackupsUsed | 0 |
| totalBackupGigabytesUsed | 0 |
+--------------------------+-------+
The quotas and so the limitations are bind to projects and not to users, so I don't know if it is possible to get a relationshit by users. The only idea I would have, would a simple bash-script, which iterates over all instances and volumes of a project and collect the information of each ressource by the user, who created it.
Found a better solution now, which also allows to get the resource usage per user of a project. It comes with the new placement-component with the stein-release of openstack (tested in train-release of openstack).
Installation of the openstack-client extension: pip install osc-placement
Ressource-usage of a project:
openstack resource usage show --os-placement-api-version 1.9 <PROJECT_ID>
Ressource-usage of a specific user within a project:
openstack resource usage show --os-placement-api-version 1.9 --user-id <USER_ID> <PROJECT_ID>
Example:
openstack resource usage show --os-placement-api-version 1.9 --user-id 98378bd3cdd94218bf7b6ef4ec80e74a 7733616a513444c2a106243db318b0dd
+----------------+-------+
| resource_class | usage |
+----------------+-------+
| VCPU | 3 |
| MEMORY_MB | 768 |
| DISK_GB | 9 |
+----------------+-------+