Disk I/O has been exceeding threshold on at least one host

Description

This alert is triggered when Disk I/O has exceeded the defined threshold for at least 15 minutes on at least one host.

Resolution Guidance

Impact When Active

When disk I/O utilization exceeds the threshold for an extended period on a host, the server's performance may be significantly impacted. This can result in slower read/write operations, application timeouts, delayed data processing, and potential service degradation for applications relying on disk operations.

To resolve this alert, you can try below steps:

• Login to Kibana.
• Click on the alert link in Relativity to navigate to Kibana dashboard.
• Identify the affected host and disk from the navigated dashboard.
• Verify for disk-intensive operations such as backups, data migrations, or large file transfers.
• Investigate the processes or applications causing high disk I/O on the affected host.
• Review recent changes or workload increases that may have caused the disk I/O spike.
• If high disk I/O persists, consider optimizing disk performance, adding storage capacity, or redistributing workload.
• Monitor the host to ensure disk I/O utilization returns to acceptable levels.

Alert Details

Alert Condition Details

Name	Value
Rule Type	Metric threshold
Group	Average
Threshold	> 0.9
Time Window	15 min
Frequency	1 min
Group alerts by	host.name

Alert Metric Details

Metric Name: system.filesystem.utilization

Metric Description: Alert triggers on Disk I/O has exceeded 90% for at least 15 minutes on at least one host.

Metric Attributes:

Attribute Name	Description
labels.device	The filesystem device name. For Windows based OS's, this is normally the primary OS drive letter.
labels.mode	Mountpoint mode such "ro", "rw", etc.
labels.mountpoint	Mountpoint path.
labels.type	Filesystem type, such as, "NTFS", "CDFS", etc.