Modifying your configuration after the upgrade

Before resuming StorageGRID operations with the new software version, you must re-enable notifications and make any other configuration changes that are required for the new release.

Before you begin

About this task

After the upgrade is complete, you must re-enable system-wide alarm notifications and event-triggered AutoSupport messages. You might also need to make configuration changes to support the new, updated, or removed features in the StorageGRID 11.4 release.

Procedure

  1. If you suppressed all system notifications during the upgrade, re-enable them.
    Even after you transition to the new alert system, you must ensure that event-triggered AutoSupport messages continue to be sent when issues arise.
    1. Select Configuration > Display Options.
    2. Unselect the Notification Suppress All check box.
    3. Click Apply Changes.
  2. Start transitioning from the legacy alarm system to the new alert system:
    1. View the count of Current Alerts on the Dashboard, and click the link to see the alerts on the Nodes page.

      Dashboard Health Panel Current Alerts
    2. Set up email notifications for alerts.
      Select Alerts > Email Setup. For details, see the instructions for monitoring and troubleshooting StorageGRID.
      Note: Because alerts and alarms are independent systems, the email setup used for alarm and AutoSupport notifications is not used for alert notifications. However, you can use the same email server for all notifications.
    3. Optionally, edit default alert rules, or create custom alert rules to meet specific monitoring requirements.
      Select Alerts > Rules. For details, see the instructions for monitoring and troubleshooting StorageGRID.
      Note: If you have created Global Custom alarms, consider creating equivalent custom alert rules using Prometheus metrics.
    4. Optionally, view the count of Legacy alarms on the Dashboard, and click the link to acknowledge or disable specific alarms.

      Dashboard Health Panel Legacy Alarms
    5. When you have fully configured alerts and alert email notifications, optionally, suppress email notifications for each mailing list you configured for the legacy alarm system.
      Suppressing notifications for a mailing list prevents recipients in that list from receiving duplicate email notifications about alarms.

      Select Support. Then, in the Alarms (legacy) section of the menu, select Email Setup. For details, see the instructions for monitoring and troubleshooting StorageGRID.

  3. If you edited any of the default alert rules or you created custom alert rules in StorageGRID 11.3, review alert rule settings to ensure these rules do not use old metric names.
    Note: During the upgrade, unedited default rules were updated automatically with the new metric names.
    1. Select Alerts > Rules.
      The Alert Rules page appears.
    2. Review the Type column to locate edited default alert rules (Default*) and custom alert rules (Custom).
    3. For each edited default alert rule or custom alert rule, review the metric names displayed in the Conditions column and determine if any rules use one of the original 11.3 metric names.
      11.3 metric name 11.4 metric name
      node_memory_MemAvailable node_memory_MemAvailable_bytes
      node_memory_MemTotal node_memory_MemTotal_bytes
      node_network_transmit_errs node_network_transmit_errs_total
      node_network_receive_errs node_network_receive_errs_total
      node_filesystem_avail node_filesystem_avail_bytes
      node_memory_MemFree node_memory_MemFree_bytes
      node_memory_Cached node_memory_Cached_bytes
      node_memory_Buffers node_memory_Buffers_bytes
      node_cpu node_cpu_seconds_total
      node_context_switches node_context_switches_total
      node_disk_io_time_ms* node_disk_io_time_seconds_total*
      node_disk_reads_completed node_disk_reads_completed_total
      node_disk_writes_completed node_disk_writes_completed_total
      node_disk_io_time_weighted node_disk_io_time_weighted_seconds_total
      node_disk_sectors_read node_disk_read_bytes_total
      node_disk_sectors_written node_disk_written_bytes_total
      node_boot_time node_boot_time_seconds
      * The metric node_disk_io_time_ms is reported in milliseconds in StorageGRID 11.3. After the upgrade to StorageGRID 11.4, the name changes to node_disk_io_time_seconds_total and values are reported in seconds. To compare values pre-upgrade to values post-upgrade, you must multiply the pre-upgrade values by 1000.
    4. If an edited or custom alert rule uses one of the old metric names, select the rule, and click Edit Rule.
    5. In the Conditions section, click the three dots to the right of the modified condition to return the metric for a default rule back to its default state. Or, update the 11.3 metric name for a custom rule to the 11.4 metric name.
      Note: When updating alert rules that use the node_disk_io_time_ms metric, you must update the trigger value for the alert as well as the metric name. The new metric is reported in seconds rather than in milliseconds.
  4. Optionally, if you want to view charts showing historical data for the renamed metrics, select Support. Then, in the Tools section of the menu, select Metrics.
    The charts on the Nodes page that are based on the renamed Prometheus metrics do not show data from before the upgrade. Historical data is only available from Support > Metrics.
    1. On the Metrics page, go to the Grafana section, and select Renamed Metrics.
    2. From the hosts menu, select the grid node whose historical data you want to view.
    3. On the Metrics page, go to the Prometheus section, and click the URL to access the Prometheus interface.
    4. Construct a query or chart using the old metric name.

    Because Prometheus metrics data is only preserved for 31 days, the Renamed Metrics dashboard and the Prometheus interface only show historical data for 31 days after the upgrade is complete. After that time, all available data for the updated metrics is shown on the charts and dashboards that use the new names of the metrics.

  5. If you used bookmarked or embedded Grafana charts to monitor StorageGRID 11.3, update the URLs of those charts.
    The URLs for the Grafana metrics available at Support > Metrics changed during the upgrade to StorageGRID 11.4:
    • The element /dashboard changed to /d
    • The element /dashboard-solo changed to /d-solo
    For example, if you bookmarked the Audit Overview charts, change the URL of the bookmark as shown below:
    • 11.3: https://sg.example.com/dashboard/audit-overview/audit-overview
    • 11.4: https://sg.example.com/d/audit-overview/audit-overview