Performance data that can be sent to an external server
Suggest changes
PDFs
Unified Manager collects a variety of performance data from all the clusters that it is monitoring. You can send specific groups of data to an external server.
Depending on the performance data that you want to chart, you can choose to send one of the following groups of statistics:
| Statistics group | Data included | Details |
|---|---|---|
Performance Monitor |
High-level performance statistics for the following objects:
|
This group provides total IOPS or latency for all LUNs and volumes in all monitored clusters. This group provides the smallest number of statistics. |
Resource Utilization |
Resource utilization statistics for the following objects:
|
This group provides utilization statistics for the node and aggregate physical resources in all monitored clusters. It also provides the statistics collected in the Performance Monitor group. |
Drill Down |
Low-level read/write and per-protocol statistics for all tracked objects:
|
This group provides read/write and per-protocol breakdowns for all seven tracked object types in all monitored clusters. It also provides the statistics collected in the Performance Monitor group and in the Resource Utilization group. This group provides the largest number of statistics. |
|
If the name of a cluster, or cluster object, is changed on the storage system, both the old and the new objects will contain performance data on the external server (called the “metric_path”). The two objects are not correlated as the same object. For example, if you change the name of a volume from “volume1_acct” to “acct_vol1”, you will see old performance data for the old volume, and new performance data for the new volume. |
See the Knowledge Base article for the list of all performance counters that can be sent to an external data provider.