Hi Community,
We have a 7 master nodes cluster. Each node holds close to 28GB maxmemory and runs on 4 cpu ubuntu azure vm.
After adding the 7th node, we started the live resharding process.
It took more than 12 hours to move 200 slots, while we need to move close to 2k slots.
It seems to be too slow for production. Worst than that, the process degrades performace of the overall cluster while it is executing.
Is there anyway to do this faster?
We tried to disable bgsave before starting resharding but it did not help.
This issue is related to #7345
Comment From: javierpajaro
For the sake of the community.
1) We sped up the resharding proccess avoinding the use of swap. 2) We reduced latency using unlink instead of delete and running bgsave by night.