Using --no-lock as we were seeing lots of flush table read locks.
What can be the highest value for --parallel switch? I would this is based on number of cores but I’ve gone upto like 200 and it still works.
Also what else can I use here to make the backup even faster?
–no-lock option disables FLUSH TABLES WITH READ LOCK. Is all your tables are InnoDB ? Make sure when using --no-lock option you should execute any DDL during backup in progress or any updates on non-innodb tables. Did you checked “backup locks” this block updates to non-transactional tables and DDL statements for all tables.
To accelerate backups you may use compression along streaming. Check details in documentation Accelerating the backup process