topic Re: Is it possible to manually backup a Cassandra node? in Dynatrace Managed Q&A

Is it possible to manually backup a Cassandra node?

GabrielPrioli — Tue, 14 Feb 2023 14:21:19 GMT

Hello forum,

Suppose a customer is running a Dynatrace Managed cluster with X amount of space for Cassandra, but they want to replicate this to a different area where they would have more space, or for backing up.

1. is this possible? is the cassandra node accessible?

2. would they be able to access the data without the cassandra node being inside a cluster node?

Re: Is it possible for a customer to manually backup a cassandra node?

Radoslaw_Szulgo — Tue, 27 Feb 2018 10:32:36 GMT

Hi,

Typically, in case an administrator wants to manage disk space effectively and increase the data store size, it is recommended to mount a bigger volume/partition under the same path that was configured as a data store. Before they need to stop the node, mount data store, and start it again.

Backups are a separate story, the path is configurable in Cluster Management Console settings.

Re 1 & 2: Cassandra is always installed on the cluster node.

Does it answers your questions?

Re: Is it possible to manually backup a Cassandra node?

sandi_pusghiyan — Thu, 26 Oct 2023 05:59:25 GMT

Hi @Radoslaw_Szulgo

i have same issue with data store size, currently dats store on our customer almost full, any idea to decrease data on mount point (/data ) so we have more space on mount point (/data)? Or indeed, increasing the size of the mount point (/data) is the only way?

Now we have activated the auto backup procedure. After the backup data is created, can we do a manual backup by moving the backup file to another disk? so the current size of the mount point (/data) can have quite a large space.

If the procedure above can be carried out, is there an impact on Dynatrace managed or data that has been backed up?

Re: Is it possible to manually backup a Cassandra node?

Radoslaw_Szulgo — Thu, 26 Oct 2023 19:23:24 GMT

Hi @sandi_pusghiyan ,

There's no good way to free up the disk space for Cassandra. Not a solution for the quick-win. If you can't mount a bigger volume, then I suggest thinking about how to split the load to the other cluster or disable gathering some metrics that are not important to you - for example, disks monitoring?

For the backup question:

If the backup was already completed and stored on the NFS, then you can copy it to another place, safely. Hope you haven't mounted your NFS to /data path - so same as your Cassandra storage.