Modify the config.yml and set the credentials
#cd /root/
#cp spark/scripts/config/config-template.yml spark/scripts/config/config.yml
Edit the srebuildd config file accordingly
Add a chord driver needed to allow the script to delete the ARC orphan keys
Copy the ring_driver0 to ring_driver1
},
"ring_driver:0": {
"alias": "arcdata",
"bstraplist": "178.33.63.219:4249",
"get_reconstruct_buffer_size": "1048576",
"ring": "DATA",
"type": "arcdata"
},
},
"ring_driver:1": {
"alias": "chord",
"bstraplist": "178.33.63.219:4249",
"ring": "DATA",
"type": "chord"
},
Restart the srebuildd connector
systemctl restart scality-srebuildd
Set the correct internal SUP password and well as the supervisor IP if the script is entented to run on a remote location
sup:
url: "https://127.0.0.1:2443"
login: "root"
password: ""
srebuildd_ip: "127.0.0.1"
Create the local directory to save the listkeys as well as the outputs for the scripts
Run a full listkeys of the DATA RING
#/root/spark_env/bin/python /root/spark/scripts/listkey.py DATA
Make sure all the listkeys are present ( you should get as many files as ring nodes)
#ls /fs/spark/listkeys-DATA.csv/*
Run the check orphan script
#/root/spark_env/bin/python /root/spark/scripts/orphan/check_orphan.py DATA
Print the list of potential orphan Keys
#cat /fs/output/output-spark-ARCORPHAN-DATA.csv/*
Report the list of real corrupted ARC keys
#/root/spark_env/bin/python /root/spark/scripts/orphan/check_orphan_corrupted.py DATA
Print the list of corrupted orphan Keys
#cat /fs/output/output-spark-ARCORPHAN-CORRUPTED-DATA.csv/*
Report the list of real corrupted ARC keys extra check for the key 70
#/root/spark_env/bin/python /root/spark/scripts/orphan/check_orphan_corrupted_single.py DATA
Print the list of corrupted orphan Keys
#cat /fs/output/output-spark-ARCORPHAN-CORRUPTED-BUT-OK-LAST-DATA.csv/*
Remove the orphans Don't run it just yet I still need to put some safe-guard before
#/root/spark_env/bin/python /root/spark/scripts/orphan/remove_orphans.py DATA