Dynatrace Managed stores multiple types of monitoring data that vary depending on the use case. There are default storage locations, as listed in Dynatrace Managed hardware and system requirements.
You might need to change existing paths for the storage locations if:
Back up your data.
An error in this procedure can result in data loss. Back up all the storage files to a safe location before you proceed and, if possible, run this procedure on a test deployment first.
Stop all Dynatrace services on the node.
By default, the script is located at <PRODUCT_PATH>/launcher/. Verify that the dynatrace.sh script has executable permissions. After you run the script, wait for it to finish and verify that no Dynatrace services remain active.
[root@host]# <PRODUCT_PATH>/launcher/dynatrace.sh stop
Move the data store to the new location.
Keep in mind that you cannot nest datastores within each other. For example, Cassandra storage cannot be a subdirectory of session storage.
[root@host]# cp -pR /old_location/cassandra/* /new_location/cassandra
Ensure that the user dynatrace:dynatrace is the owner of this new directory.
[root@host]# chown -R dynatrace:dynatrace /new_location
Update the new data location in /etc/dynatrace.conf.
The location must be either an absolute path or a value based on predefined variables. It must be a directory, and it cannot be a symlink.
Update the following section:
# Paths to directories with component's dataDATASTORE_PATH = /var/opt/dynatrace-managedCASSANDRA_DATASTORE_PATH = DATASTORE_PATH/CASSANDRA_DIRELASTICSEARCH_DATASTORE_PATH = DATASTORE_PATH/ELASTICSEARCH_DIRSERVER_DATASTORE_PATH = DATASTORE_PATH/SERVER_DIRSERVER_REPLAY_DATASTORE_PATH = SERVER_DATASTORE_PATH/replayDataNODEKEEPER_DATASTORE_PATH = DATASTORE_PATH/NODEKEEPER_DIR
Run the reconfigure via the installer. Use the nohup command to prevent interruption of script execution (such as session disconnect) during important operations.
[root@host]# nohup <PRODUCT_PATH>/installer/reconfigure.sh --no-start &
This step is crucial to propagating the changes from /etc/dynatrace.conf to all relevant configuration files. The --no-start flag will allow you to make a final verification of the changes before starting any of the Dynatrace services.
The script output should look similar to the following:
Reconfiguration completed successfully after 1 minute 9 seconds.Dynatrace binaries are located in directory /opt/dynatrace-managedDynatrace data is located in directory /new_locationDynatrace metrics repository is located in directory /new_location/cassandraDynatrace Elasticsearch store is located in directory /new_location/elasticsearchDynatrace server store is located in directory /new_location/sessionstorageDynatrace session replay store is located in directory /new_location/replayDataDon't forget to start Dynatrace Server and log in at https://<your_ip>
Start all Dynatrace services.
[root@host]# <PRODUCT_PATH>/launcher/dynatrace.sh start
Check the logs to make sure all services started without issues.
The directory paths included in the following table are the default paths. Actual paths might vary if you've installed to a custom directory.
If you customized the storage locations, place SERVER_DATASTORE_PATH, CASSANDRA_DATASTORE_PATH, and ELASTICSEARCH_DATASTORE_PATH in separate directories, and make none of them a subdirectory of another.
| Directory symbol | Directory path | Description | Required free disk space for installing | Required free disk space for upgrading |
|---|---|---|---|---|
|
| Main directory for self-monitoring OneAgent binaries | 3 GB 1 | 1.8 GB 1 |
|
| Main directory for Dynatrace Managed binaries | 12 GB 1 | 10 GB 1 |
|
| Main directory for Dynatrace Managed data | 24 GB | 3 GB |
|
| Logs of all Dynatrace Managed components, services, and tools | 2 GB | 1 GB |
Not configurable (fixed path) |
| Logs of the installer, upgrade, and configurator, such as | 1 MB 2 | 1 MB 2 |
|
| Metrics repository | 25 GB | 1 GB |
|
| Elasticsearch store | 3 GB | 1 GB |
|
| Transactions store | 14 GB | 1 GB |
|
| Session Replay store | 14 GB | 1 GB |
|
| OneAgent and ActiveGate installation packages (if downloaded by Dynatrace Server or installed from standalone packages) | 20 GB | 1 GB |
|
| Dynatrace Managed installer for adding nodes to a Managed Cluster, prepared during installation or upgrade | 2 GB | 1 GB |
This value is part of a total value required for the /opt directory. Total required free disk space is the sum of the required disk space for /opt/dynatrace-managed and /opt/dynatrace.
The installer verifies the free space on the filesystem that holds /var/log. Requirements for paths located on the same filesystem are added together.
The installer writes its own logs to /var/log/dynatrace, a fixed path that cannot be changed and that OneAgent shares through the /var/log/dynatrace/oneagent subdirectory. If /var/log/dynatrace is a symbolic link or carries restrictive POSIX ACLs, verify that the Dynatrace product user has write access to the effective target directory, including the ACL mask.
Nested mount points are not supported in the DATASTORE_PATH/server location for /tenantData or /replayData.
To extend the size of the storage volume, you need to follow the disk extension procedure from your storage provider. If you use dynamic storage, you most likely can adjust the disk parameters (size, bandwidth, type) and the host will automatically apply it. Otherwise, a disk re-mounting might be necessary. If you've increased the size of your volume, then you must also extend the volume's partition to use the additional storage capacity. Dynatrace Managed services should automatically discover any changes to this.
If there's insufficient disk space on any of your used mount points, the Managed Cluster generates a Managed Cluster event like this:
You need at least <required MB> free for <storage type>. Currently, only <current free disk space> of a total of <disk size> is available.
Providing enough disk space is vital to the health of your Managed Cluster and ensures your monitoring data is safe, consistent, and available. Ignoring that message might cause data loss.
If the transaction data storage is affected, you'll see an additional Managed Cluster event that automatically truncates the data and adapts the data retention period. To read more, see Adaptive data retention.