Backup and restore with Google Cloud Storage

Use Medusa to backup and restore Apache Cassandra® data in Kubernetes to Google Cloud Storage (GCS).

Medusa is a Cassandra backup and restore tool. It’s packaged with K8ssandra and supports a variety of backends, including GCS for storage in Google Kubernetes Engine (GKE) environments.

Introduction

Google Cloud Storage (GCS) is a RESTful online file storage web service for storing and accessing data on Google Cloud Platform (GCP) / GKE infrastructure. The service combines the performance and scalability of Google’s cloud with advanced security and sharing capabilities.

For details about GCS, see the Google Cloud Storage documentation.

Create a role for backups

In order to perform backups in GCS, Medusa needs to use a service account with appropriate permissions. See this permissions setup article.

Using the Google Cloud SDK, run the following command to create the MedusaStorageRole (set the $GCP_PROJECT env variable appropriately):

gcloud iam roles create MedusaStorageRole \
        --project ${GCP_PROJECT} \
        --stage GA \
        --title MedusaStorageRole \
        --description "Custom role for Medusa for accessing GCS safely" \
        --permissions storage.buckets.get,storage.buckets.getIamPolicy,storage.objects.create,storage.objects.delete,storage.objects.get,storage.objects.getIamPolicy,storage.objects.list

Create a GCS bucket

Create a bucket for each Cassandra cluster, using the following command line (set the env variables appropriately):

gsutil mb -p ${GCP_PROJECT} -c regional -l ${LOCATION} ${BUCKET_URL}

Create a service account and download its keys

Medusa will require a credentials.json file with the informations and keys for a service account with the appropriate role in order to interact with the bucket.

Create the service account (if it doesn’t exist yet):

gcloud --project ${GCP_PROJECT} iam service-accounts create ${SERVICE_ACCOUNT_NAME} --display-name ${SERVICE_ACCOUNT_NAME}

Configure the service account with the role

Once the service account has been created, and considering jq is installed, run the following command to add the MedusaStorageRole to it, for our backup bucket:

gsutil iam set <(gsutil iam get ${BUCKET_URL} | jq ".bindings += [{\"members\":[\"serviceAccount:${SERVICE_ACCOUNT_NAME}@${GCP_PROJECT}.iam.gserviceaccount.com\"],\"role\":\"projects/${GCP_PROJECT}/roles/MedusaStorageRole\"}]") ${BUCKET_URL}

Configure Medusa

Generate a json key file called credentials.json, for the service account:

gcloud --project ${GCP_PROJECT} iam service-accounts keys create credentials.json --iam-account=${SERVICE_ACCOUNT_NAME}@${GCP_PROJECT}.iam.gserviceaccount.com

Place this file on all Cassandra nodes running medusa under /etc/medusa and set the rights appropriately so that only users running Medusa can read/modify it. Set the key_file value in the [storage] section of /etc/medusa/medusa.ini to the credentials file:

bucket_name = my_gcs_bucket
key_file = /etc/medusa/credentials.json

Medusa should now be able to access the bucket and perform all required operations, as explained below.

Deploy K8ssandra

Now that you have GCS set up, install K8ssandra, which includes Medusa:

helm install k8ssandra k8ssandra/k8ssandra -n k8ssandra

Now wait for the Cassandra cluster to be ready by using the following wait command:

kubectl wait --for=condition=Ready cassandradatacenter/dc1 --timeout=900s -n k8ssandra

When ready, you should now see a list of pods. Example:

kubectl get pods -n k8ssandra

Output:

NAME                                                  READY   STATUS      RESTARTS   AGE
k8ssandra-cass-operator-547845459-dwg68               1/1     Running     0          6m36s
k8ssandra-dc1-default-sts-0                           3/3     Running     0          5m56s
k8ssandra-dc1-stargate-776f88f945-p9twg               0/1     Running     0          6m36s
k8ssandra-grafana-75b9cb64cc-kndtc                    2/2     Running     0          6m36s
k8ssandra-kube-prometheus-operator-5bdd97c666-qz5vv   1/1     Running     0          6m36s
k8ssandra-medusa-operator-d766d5b66-wjt7j             1/1     Running     0          6m36s
k8ssandra-reaper-5f9bbfc989-j59xk                     1/1     Running     0          2m48s
k8ssandra-reaper-operator-858cd89bdd-7gfjj            1/1     Running     0          6m36s
k8ssandra-reaper-schema-4gshj                         0/1     Completed   0          3m3s
prometheus-k8ssandra-kube-prometheus-prometheus-0     2/2     Running     1          6m32s

Create some data and back it up

Next, let’s define some sample data in Cassandra by creating a test_data.cql file that contains DDL and DML statements:

CREATE KEYSPACE medusa_test  WITH replication = {'class': 'SimpleStrategy', 'replication_factor': 1};
USE medusa_test;
CREATE TABLE users (email text primary key, name text, state text);
insert into users (email, name, state) values ('[email protected]', 'Alice Smith', 'TX');
insert into users (email, name, state) values ('[email protected]', 'Bob Jones', 'VA');
insert into users (email, name, state) values ('[email protected]', 'Carol Jackson', 'CA');
insert into users (email, name, state) values ('[email protected]', 'David Yang', 'NV');

Copy the CQL file into the Cassandra pod; that is, the StatefulSet one, which contains -sts- in its name:

kubectl cp test_data.cql k8ssandra-dc1-default-sts-0:/tmp -n k8ssandra -c cassandra

Now extract the password to access Cassandra with the k8ssandra-superuser. (The password is different for each installation unless it is explicitly set at install time.)

kubectl get secret k8ssandra-superuser -n k8ssandra -o jsonpath="{.data.password}" | base64 --decode ; echo

Output:

XHsZ943WBg5RPNhVAT8x

Let’s now run the uploaded cql script and check that you can read the data.

kubectl exec -it k8ssandra-dc1-default-sts-0 -n k8ssandra -c cassandra -- cqlsh -u k8ssandra-superuser -p XHsZ943WBg5RPNhVAT8x -f /tmp/test_data.cql
kubectl exec -it k8ssandra-dc1-default-sts-0 -n k8ssandra -c cassandra -- cqlsh -u k8ssandra-superuser -p XHsZ943WBg5RPNhVAT8x -e "SELECT * FROM medusa_test.users"

Output:

 email             | name          | state
-------------------+---------------+-------
 [email protected] |   Alice Smith |    TX
   [email protected] |     Bob Jones |    VA
 [email protected] |    David Yang |    NV
 [email protected] | Carol Jackson |    CA

(4 rows)

Now backup this data, and check that files get created in your GCS bucket.

helm install my-backup k8ssandra/backup -n k8ssandra --set name=backup1,cassandraDatacenter.name=dc1

Because the backup operation is asynchronous, you can monitor its completion by running the following command:

kubectl get cassandrabackup backup1 -n k8ssandra -o jsonpath={.status.finishTime}

As long as this command doesn’t output a date and time, you know that the backup is still running. With the amount of data present and the fact that you’re using a locally accessible backend, this should complete quickly.

Now refresh the GCS UI and you should see some files in the k8ssandra-medusa bucket.

In the GCS UI, you should see an index folder, which is the Medusa backup index, and another folder that is specific to each Cassandra node in the cluster.

Deleting the data and restoring the backup

Now delete the data by truncating the table, and check that the table is empty.

kubectl exec -it k8ssandra-dc1-default-sts-0 -n k8ssandra -c cassandra -- cqlsh -u k8ssandra-superuser -p XHsZ943WBg5RPNhVAT8x -e "TRUNCATE medusa_test.users"
 kubectl exec -it k8ssandra-dc1-default-sts-0 -n k8ssandra -c cassandra -- cqlsh -u k8ssandra-superuser -p XHsZ943WBg5RPNhVAT8x -e "SELECT * FROM medusa_test.users"

Output:

 email | name | state
-------+------+-------

(0 rows)

Now restore the backup taken previously:

helm install restore-test k8ssandra/restore --set name=restore-backup1,backup.name=backup1,cassandraDatacenter.name=dc1 -n k8ssandra

The restore operation will take a little longer because it requires K8ssandra to stop the StatefulSet pod and perform the restore as part of the init containers, before the Cassandra container can start. You can monitor progress using this command:

watch -d kubectl get cassandrarestore restore-backup1 -o jsonpath={.status} -n k8ssandra

The restore operation is fully completed once the finishTime value appears in the output. Example:

{"finishTime":"2021-03-30T13:58:36Z","restoreKey":"83977399-44dd-4752-b4c4-407273f0339e","startTime":"2021-03-30T13:55:35Z"}

Verify that you can now read the restored data from the previously truncated table:

kubectl exec -it k8ssandra-dc1-default-sts-0 -n k8ssandra -c cassandra -- cqlsh -u k8ssandra-superuser -p XHsZ943WBg5RPNhVAT8x -e "SELECT * FROM medusa_test.users"

Output:

 email             | name          | state
-------------------+---------------+-------
 [email protected] |   Alice Smith |    TX
   [email protected] |     Bob Jones |    VA
 [email protected] |    David Yang |    NV
 [email protected] | Carol Jackson |    CA

(4 rows)

Success! You’ve successfully restored your lost data in just a few commands.

Next steps

See the following reference topics:


Last modified July 22, 2021: Fix links format (#999) (aac72ff)