Issues tables

When you deploy version 25.1 on newer RHEL/kernel environments, the readiness probe script /scripts/ping_readiness_local.sh triggers zombie processes in the redis-cluster stateful set pods.

During an upgrade from version 24.3 to 24.4, the upgrade-configuration.sh script failed to preserve the custom LOG_ES_PASSWD value provided during the earlier deployment. Instead, the default password (Es_L0g#p@SS) was retained in the secrets.txt file after running the script, indicating the script did not correctly carry forward the configured value.

When deploying Helix Common Services version 24.3, the installer attempted to pull the global-config-service image from the lp0mz repository, but the image was available in the lp0lz repository, resulting in an ImagePullError during installation.

BMC Helix IT Operations Management

When upgrading to version 25.1, the application services might enter a CrashLoopBackOff state due to a Redis-cluster replica error.
This issue occurs because the Redis cluster fails to form the expected number of replicas, resulting in a NOREPLICAS error during write operations.

The issue is resolved by automating the recovery process through a job container.

BMC Helix AIOps

 The BMC Helix AIOps service creation fails with the permission denied error.

When accessing the AIOps Service or Situation page, the UI might hang or fail to load completely.

This occurs because of intermittent issues in the UI rendering process, which can cause the page to remain blank or get stuck in a loading state, affecting access to AIOps services and situation analysis.

Workaround:

Contact the BMC Helix support team for further assistance in resolving this issue.

In OnPrem stack deployments with version 25.2.00, even if the BHCO (BMC Helix Capacity Optimization) product is disabled (i.e., Optimize="No" in the deployment.config file or via installer options), the related infrastructure components are still deployed. This includes BHCO-related Victoria Metrics Aggregate services such as:

• victoria-metrics-aggregate-vmstorage-agg
• victoria-metrics-aggregate-vmselect-agg
• victoria-metrics-aggregate-vminsert-agg

This results in unnecessary resource allocation, potentially affecting sizing and overall system performance.

Workaround:

After deployment, manually uninstall the BHCO-related infrastructure components to avoid unnecessary resource consumption. This can be done by:

1. Identifying and deleting the BHCO-related Helm charts using Helm or kubectl
   helm uninstall victoria-metrics-aggregate -n <namespace> 
2. Clear the persistent volume storage:
   kubectl delete pvc <associated vmstorage-volume-victoria-metrics-aggregate-vmstorage-agg>

When using tctl version 6.1.4 or later, the following error might occur during execution:

tls: handshake failure 

This happens because the updated Go runtime enforces stricter TLS security settings, which can conflict with older TLS configurations on the server.

Workaround:
Before running tctl, set the following environment variable to restore compatibility with older TLS settings:

export GODEBUG=tlsrsakex=1 

This enables RSA-based key exchange, allowing tctl to connect successfully to servers with legacy TLS setups.

In on-premises deployments, MinIO may fail the SSL handshake if using a self-signed certificate.

This typically results in errors when accessing or uploading files to MinIO over HTTPS.

Workaround:
Add the MinIO self-signed certificate to the trusted certificate store. For Java-based services, use the following command:

keytool -import -alias minio-cert -keystore $JAVA_HOME/lib/security/cacerts -file /path/to/minio.crt