As part of the verification testing of Machine Learning algorithms, we are requesting customer test data be collected so that real world examples of metric data processing can be evaluated in partnership with the customer.
Release : 2.0.0*
Component : MF OPERATIONAL INTELLIGENCE
MOI Data Extraction for Machine Learning Algorithm Testing
- A Support Case should be opened with a subject like "Customer Data Collection for Testing of Machine Learning Algorithm"
- The extract data utility in the zipped tar file attached to this knowledge doc needs to be executed on the host where your MOI Appliance is running
- MOI Appliance should be running and all containers should be healthy
- The Host system needs to have free disk space to create and tar up the data backup. Without specific Cassandra database stats, we cannot provide an exact estimate of the amount of space needed. Our best guess is in the 5 to 20 GB range.
- It is recommended that a MOI Appliance with a limited number of metric data feeds be selected for this data extract as the size of the extracted data can be voluminous.
**How to Use:**
1) Ensure the /var/opt/moi/share/backups directory is created or empty if existing already. All extracted data will be copied here before being zipped and tar'ed up.
2) Download the moi-data-extractor.tar_1616445369805.gz utility to a folder of your selection on the Linux system where the MOI data needs to be extracted.
3) Extract the content of the moi-data-extractor.tar_1616445369805.gz archive. This will create an moi-data-extractor folder
tar -xvf moi-data-extractor.tar_1616445369805.gz
4) Change working directory to "moi-data-extractor" folder
5) Run the extract-data.sh script.
6) Wait until the extract-data.sh completes. Please note that depending on the amount of data, this script can run for several hours.
7) Once the data backup is done, you should see the 'moi-data.tar.gz' in directory /var/opt/moi/share. Please upload the moi-data.tar.gz file to the Support Case that was opened for this data capture.