Insufficient disk space in Cassandra

Symptom

Your Apigee hybrid installation process fails due to insufficient disk space for the Cassandra database. The installation process fails with the following error message:

Error: UPGRADE FAILED: cannot patch "default" with kind ApigeeDatastore:
Internal error occurred

Additionally, the Cassandra log contains a warning similar to the following:

WARN  [main] 2024-06-18 12:34:55,583 DatabaseDescriptor.java:579 - Only
62.440GiB free across all data volumes. Consider adding more capacity to your
cluster or removing obsolete snapshots

Common diagnosis steps

Use the Apigee hybrid must-gather script to collect the logs from the Cassandra pods for review.

Possible cause

Cause	Description
Out of disk space	The available disk space for Cassandra is less than 50%.

Cause 1: Out of disk space

When upgrading Apigee hybrid, ensure you have more than 50% disk space available for all Cassandra pods.

Diagnosis

1. From the Apigee hybrid must-gather script, examine the output of the following command to check the total size of the volume associated with Cassandra:

   kubectl get pv -n APIGEE_NAMESPACE

   Sample Output

   NAME                                     CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM                                            STORAGECLASS   VOLUMEATTRIBUTESCLASS   REASON    AGE
   pvc-0b6b2daa-d512-4780-9021-fc97293a8154 10Gi     RWO          Delete         Bound  apigee/cassandra-data-apigee-cassandra-default-1 standard-rwo    <unset>                 -        20d
   pvc-2263fc7c-e057-406a-ad60-38573733c92d 10Gi     RWO          Delete         Bound  apigee/cassandra-data-apigee-cassandra-default-2 standard-rwo    <unset>                 -        20d
   pvc-8c854fe9-adaa-440f-90d9-d15497e7f530 10Gi     RWO          Delete         Bound  apigee/cassandra-data-apigee-cassandra-default-0 standard-rwo    <unset>                 -        20d
   

2. Check the nodetool status output to see the disk utilized by each Cassandra pod. The disk space utilization for Persistent Volume Claims (PVCs) assigned to Cassandra pods should not surpass 50% of the available storage capacity. Refer to the following command:

   kubectl -n APIGEE_NAMESPACE -c apigee-cassandra exec apigee-cassandra-default-1 -- bash -c 'nodetool -u cassandra -pw $CASS_PASSWORD status'

   Where CASS_PASSWORD is the cassandra.auth.default.password as mentioned in Configuration property reference.

   Sample Output

   Datacenter: us-west3
   ========================
   Status=Up/Down
   |/ State=Normal/Leaving/Joining/Moving
   --  Address      Load        Tokens  Owns (effective)  Host ID                               Rack
   UN  10.10.10.1   6.69 GiB   256     100.0%            2a184030-444-4155-8375-156e87539711      a
   UN  10.10.10.2   7.42 GiB   256     53.1%             f22b66c7-4444-4000-b9c0-5a71ee6315a8     c
   UJ  10.10.10.3   7.39 GiB   256     ?                 78fb2737-4444-4468-a786-e473ead115b5     b
   UN  10.10.10.4   7.81 GiB   256     47.0%             779fcdf1-4444-4186-bcbb-2844e59629c2     b
   UN  10.10.10.5   6.91 GiB   256     53.0%             55ca07b2-4444-4967-b321-6477d50f9846     b
   UN  10.10.10.6   7.41 GiB   256     46.9%             0cf33585-444-46ce-811f-9c6376ed58ac      c
   

   (Optional) Use kubectl to execute a command inside the Cassandra container to check the disk usage percentage.

   kubectl exec -it apigee-cassandra-default-0 -n apigee – df -h /opt/apigee/data
   kubectl exec -it apigee-cassandra-default-1 -n apigee – df -h /opt/apigee/data
   kubectl exec -it apigee-cassandra-default-2 -n apigee – df -h /opt/apigee/data
   

   Example

   The following example shows where disk space utilization is over 50%. This would cause problems for Apigee hybrid upgrades.

   kubectl exec -it apigee-cassandra-default-0 -n apigee -- df -h /opt/apigee/data
   Defaulted container "apigee-cassandra" out of: apigee-cassandra, apigee-cassandra-ulimit-init (init)
   Filesystem      Size    Used    Avail   Use%    Mounted on
   /dev/sdb         10G    6G      4G      60%     /opt/apigee/data
   
   
    kubectl exec -it apigee-cassandra-default-1 -n apigee -- df -h /opt/apigee/data
   Defaulted container "apigee-cassandra" out of: apigee-cassandra, apigee-cassandra-ulimit-init (init)
   Filesystem      Size    Used    Avail   Use%    Mounted on
   /dev/sdb         10G    5.9G    4.1G    59%     /opt/apigee/data
   
   
   kubectl exec -it apigee-cassandra-default-2 -n apigee -- df -h /opt/apigee/data
   Defaulted container "apigee-cassandra" out of: apigee-cassandra, apigee-cassandra-ulimit-init (init)
   Filesystem      Size    Used    Avail   Use%    Mounted on
   /dev/sdb         10G    5.6G    4.4G    56%     /opt/apigee/data
   
   

Resolution

If you find disk utilization is above 50%, scale the Cassandra disk capacity to ensure it is below 50% before performing an Apigee hybrid upgrade.

To expand disk capacity, see Manage Cassandra persistent storage size.

Must gather diagnostic information

If the problem persists even after following the above instructions, gather the following diagnostic information and then contact Google Cloud Customer Care:

• Your Google Cloud Project ID.
• Your Apigee hybrid organization.
• The overrides.yaml files from both the source and new regions (remember to mask any sensitive information).
• The outputs from the commands in Apigee hybrid must-gather.