Cluster configuration and performance data collection activity
-
PDF of this doc site
- Perform configuration and administrative tasks
- Monitor and manage storage
- Monitor and manage cluster performance
- Monitor and manage cluster health
Collection of separate PDF docs
Creating your file...
The collection interval for cluster configuration data is 15 minutes. For example, after you have added a cluster, it takes 15 minutes to display the cluster details in the Unified Manager UI. This interval applies when making changes to a cluster too.
For example, if you add two new volumes to an SVM in a cluster, you see those new objects in the UI after the next polling interval, which could be up to 15 minutes.
Unified Manager collects current performance statistics from all monitored clusters every five minutes. It analyzes this data to identify performance events and potential issues. It retains 30 days of five-minute historical performance data and 180 days of one-hour historical performance data. This enables you to view very granular performance details for the current month, and general performance trends for up to a year.
The collection polls are offset by a few minutes so that data from every cluster is not sent at the same time, which could affect performance.
The following table describes the collection activities that Unified Manager performs:
Activity | Time interval | Description |
---|---|---|
Performance statistics poll |
Every 5 minutes |
Collects real-time performance data from each cluster. |
Statistical analysis |
Every 5 minutes |
After every statistics poll, Unified Manager compares the collected data against user-defined, system-defined, and dynamic thresholds. If any performance thresholds have been breached, Unified Manager generates events and sends email to specified users, if configured to do so. |
Configuration poll |
Every 15 minutes |
Collects detailed inventory information from each cluster to identify all the storage objects (nodes, SVMs, volumes, and so on). |
Summarization |
Every hour |
Summarizes the latest 12 five-minute performance data collections into hourly averages. The hourly average values are used in some of the UI pages, and they are retained for 180 days. |
Forecast analysis and data pruning |
Every day after midnight |
Analyzes cluster data to establish dynamic thresholds for volume latency and IOPS for the next 24 hours. Deletes from the database any five-minute performance data older than 30 days. |
Data pruning |
Every day after 2 a.m. |
Deletes from the database any events older than 180 days and dynamic thresholds older than 180 days. |
Data pruning |
Every day after 3:30 a.m. |
Deletes from the database any one-hour performance data older than 180 days. |