GitHub Runner pods - Listening for Jobs

Hi Team,

We are hosting Github actions runner in a self-hosted mode in our AWS EKS cluster, sometimes I can see the runner pods are just waiting idle for more than 4-5 hours. Queued jobs are assigned to other runner pods and these pods are just idle and consuming overall cluster resources and eventually we have manually clear these idle pods.

Runner version - 2.335.1
ARC version - 0.14.2

[RUNNER 2026-06-25 11:59:19Z INFO RunnerServer] EstablishVssConnection
[RUNNER 2026-06-25 11:59:19Z INFO RunnerServer] Establish connection with 60 seconds timeout.
[RUNNER 2026-06-25 11:59:19Z INFO GitHubActionsService] Starting operation Location.GetConnectionData
[RUNNER 2026-06-25 11:59:19Z INFO GitHubActionsService] Finished operation Location.GetConnectionData
[RUNNER 2026-06-25 11:59:19Z INFO GitHubActionsService] Finished operation Location.GetConnectionData
[RUNNER 2026-06-25 11:59:19Z INFO GitHubActionsService] Finished operation Location.GetConnectionData
[RUNNER 2026-06-25 11:59:20Z INFO BrokerMessageListener] VssConnection created
[RUNNER 2026-06-25 11:59:20Z INFO Terminal] WRITE LINE: 

√ Connected to GitHub
[RUNNER 2026-06-25 11:59:20Z INFO Terminal] WRITE LINE: 

[RUNNER 2026-06-25 11:59:20Z INFO RSAFileKeyManager] Loading RSA key parameters from file /home/runner/.credentials_rsaparams
[RUNNER 2026-06-25 11:59:20Z INFO RSAFileKeyManager] Loading RSA key parameters from file /home/runner/.credentials_rsaparams
[RUNNER 2026-06-25 11:59:21Z INFO GitHubActionsService] AAD Correlation ID for this token request: Unknown
[RUNNER 2026-06-25 11:59:21Z INFO BrokerMessageListener] Session created.
[RUNNER 2026-06-25 11:59:21Z INFO Terminal] WRITE LINE: Current runner version: '2.335.1'
Current runner version: '2.335.1'
[RUNNER 2026-06-25 11:59:21Z INFO Terminal] WRITE LINE: 2026-06-25 11:59:21Z: Listening for Jobs
2026-06-25 11:59:21Z: Listening for Jobs
[RUNNER 2026-06-25 11:59:21Z INFO JobDispatcher] Set runner/worker IPC timeout to 30 seconds.
[RUNNER 2026-06-25 12:49:30Z INFO RSAFileKeyManager] Loading RSA key parameters from file /home/runner/.credentials_rsaparams
[RUNNER 2026-06-25 12:49:31Z INFO GitHubActionsService] AAD Correlation ID for this token request: Unknown
[RUNNER 2026-06-25 13:39:40Z INFO RSAFileKeyManager] Loading RSA key parameters from file /home/runner/.credentials_rsaparams
[RUNNER 2026-06-25 13:39:40Z INFO GitHubActionsService] AAD Correlation ID for this token request: Unknown
[RUNNER 2026-06-25 14:29:49Z INFO RSAFileKeyManager] Loading RSA key parameters from file /home/runner/.credentials_rsaparams
[RUNNER 2026-06-25 14:29:50Z INFO GitHubActionsService] AAD Correlation ID for this token request: Unknown

Start with the reported runner logs and the AWS EKS deployment using ARC 0.14.2 and runner 2.335.1. Reproduce the condition where pods remain at “Listening for Jobs” while queued jobs go elsewhere, then trace the runner and ARC behavior; done means identifying the cause and providing a verified fix or workaround for idle pods.