MySql Replication Performance on Master after replication is resumed?

You can improve the latter by slinging a crossover cable between master and slave and routing all the replication traffic over that, keeping it off your LAN If the former is the problem, I'd say your best option is to get yourself more and/or faster disks, or even SSD! You could also do your binary logging to separate disks so that it doesn't slow down query activity Overall it sounds like you've just got a capacity problem, and you need to provide sufficient resources to deal with the load you're pointing at it - no amount of juggling is going to let you work around that. I've not encountered a config where binary logging is the limiting factor.

Any basic RAID config should be handling at least 300M/sec, and for this to be a problem you need to be holding things up for more than say 30 sec at a time, implying that you need to move more than 9Gb of binlogs every night - are you really generating that much data? Another alternative is to use DRBD for replication that way the slave doesn't have to run the replicated queries or deal with binlogs at all, though it has other complications.

You can improve the latter by slinging a crossover cable between master and slave and routing all the replication traffic over that, keeping it off your LAN. If the former is the problem, I'd say your best option is to get yourself more and/or faster disks, or even SSD! You could also do your binary logging to separate disks so that it doesn't slow down query activity.

Overall it sounds like you've just got a capacity problem, and you need to provide sufficient resources to deal with the load you're pointing at it - no amount of juggling is going to let you work around that. I've not encountered a config where binary logging is the limiting factor. Any basic RAID config should be handling at least 300M/sec, and for this to be a problem you need to be holding things up for more than say 30 sec at a time, implying that you need to move more than 9Gb of binlogs every night - are you really generating that much data?

Another alternative is to use DRBD for replication - that way the slave doesn't have to run the replicated queries or deal with binlogs at all, though it has other complications.

Thanks for your reply. 8-gigs RAM on each server. The issue was iowait.

Ended up adding "SET SQL_BIN_LOG = 0" right before many of the load operations (on the master). This essentially solved the problem since I did not need to have much of the temporary load table data replicated across. The load table data was indeed substantial..multiple gigs for sure.

– Ben Jun 16 at 14:16.

I cant really gove you an answer,but what I can give you is a way to a solution, that is you have to find the anglde that you relate to or peaks your interest. A good paper is one that people get drawn into because it reaches them ln some way.As for me WW11 to me, I think of the holocaust and the effect it had on the survivors, their families and those who stood by and did nothing until it was too late.