Once the system is installed and you are familiar with the CLI and GUI, you can connect to one of the servers and try it out.
To perform a sanity check that the WEKA cluster is configured and IOs can be performed on it, do the following procedures:
.
.
.
To validate that the WEKA cluster and IT environment are best configured to benefit from the WEKA filesystem, do the following procedure:
.
A filesystem must reside in a filesystem group. Create a filesystem group:
2. Create a filesystem within that group:
For creating an additional filesystem, it is first needed to decrease the default
filesystem SSD size:
To mount a filesystem, create a mount point and call the mount command:
2. Check that the filesystem is mounted:
Write data to the filesystem:
This has completed the sanity check that the WEKA cluster is configured and IOs can be performed on it.
To ensure that the WEKA cluster and the IT environment are well configured, more complex IO patterns and benchmark tests should be conducted using the FIO utility.
Although results can vary using different servers and networking, it is not expected to be very different than what many other customers and we achieved. A properly configured WEKA cluster and IT environment should yield similar results described in the WEKA performance tests section.
The numbers achieved in the benchmark tests, as described in the WEKA performance tests section, are not just achieved in a controlled environment. Similar numbers should be achieved using a similar configuration if the WEKA cluster and IT environment are properly configured.
Related topic
In AWS installation via the , the default filesystem group and filesystem are created. The default
filesystem is created with the entire SSD capacity.
For more information about filesystems and filesystem groups, see .
In AWS installation via the , the default
filesystem is already mounted under /mnt/weka.
For more information about mounting filesystems and mount options, refer to .
If the numbers achieved in your environment significantly vary from those, contact the before running any other workload on the WEKA cluster.
This is a quick reference guide using the CLI to perform the first IO in the WEKA filesystem.