This PR improves the error handling related to when/if indexing jobs error out (for example, due to OOM errors) before having a chance to update the job status in cosmosdb.
To resolve this edge case, the index job manager (a cronjob) has been updated to check both the index job status in cosmosdb and checks whether or not there is a k8s job running.
This PR improves the error handling related to when/if indexing jobs error out (for example, due to OOM errors) before having a chance to update the job status in cosmosdb.
To resolve this edge case, the index job manager (a cronjob) has been updated to check both the index job status in cosmosdb and checks whether or not there is a k8s job running.