Feedback
b20b9d71-bfe0-4fb3-b1e4-d3288ea8ee28
Agent server is starving
Description
This alert is triggered when an agent server in Relativity is found to be starving - meaning it has not associated with any resource pool, leaving it unable to process work.
Resolution Guidance
Impact When Active
When an agent server is starving, it is not assigned to any resource pool and therefore cannot run agents or process jobs. This can result in background processing tasks - such as analytics, search index builds, and other automated operations - failing to execute or becoming severely delayed. Continued starvation may cause pipeline backlogs and impact system performance and reliability.
How To Resolve
- Log into Relativity.
- Navigate to Servers.
- Click on agent server Name.
- Under the Resource Pool section, click on Add resource pool, select Default resource pool, click on > (Move selected left to right).
- Click Apply.
- Verify that the server is now active and agents are being scheduled on it.
Additional Checks
- Confirm that the resource pool you are assigning the server to has active agents configured.
- If the server remains in a starving state after pool assignment, check the Relativity Service Host Manager on the agent server and restart it if necessary.
- Review the Agents tab to ensure agents are enabled and distributing work to the server.
Alert Details
Alert Condition Details
| Name | Value |
|---|---|
| Rule Type | Elastic Query |
| Data View | metrics-* |
| Filter Query | relsvr.agent.server.starving:1 |
| Group | Count |
| Threshold | > 0 |
| Time Window | 30 sec |
| Frequency | 30 sec |
Alert Metric Details
Metric Name: relsvr.agent.server.starving
Metric Description: Alert triggers when the agent server starving count is greater than 0 for the last 30 seconds, indicating the server has no resource pool assignment and cannot process agent jobs.
Metric Attributes:
| Attribute Name | Description |
|---|---|
| relsvr.agent.server.starving | 1 |
| labels.Application | Environment Administration & Operations |
| labels.CorrelationId | (empty) |
| labels.HostName | (empty) |
| labels.IsHealthy | false |
| labels.MeasureTypeName | HealthCheck |
| labels.MeasureType | 3 |
| labels.Name | Agent Server Starving |
| labels.System | Resource Servers |
| labels.SubSystem | Agents |
| labels.TimeStamp | Date and Time |
| labels.WorkspaceGuid | (empty) |
| labels.Exception.Message | (empty) |
| labels.Exception.InnerExceptionMessage | (empty) |
| labels.Exception.LoggingCorrelationGuid | (empty) |
| labels.Message | Message describes the issue |
