Issue 1: Too many pilot replication delays from the library
Workaround: Reduce the From library to three to five
Issue 2: Hardware configuration of Master and slave machine is inconsistent, slave hardware configuration is lower than main hardware configuration
WORKAROUND: Increase the hardware configuration of the slave machine to match the hardware configuration of the master
Issue 3: Too many statements for slow SQL
Workaround: Refine the slow query statement
Issue 4: Master-slave replication design Issues
Workaround: Master-slave replication single-threaded, if the main library write concurrency too large, too late to transfer to the slave library, will cause delay, using a higher version of the database with multithreading to copy
Issue 5: Network latency between master and slave libraries
Workaround: Increase network bandwidth
Issue 6: High reading and writing pressure in the main library
Workaround: Add buffer and cache to the front-end of the schema
Note: Gao said no matter how much delay, as long as not affect the business is OK
The Master never synchronizes:
First hardware and network and master-slave single-threaded copy is we can not immediately change, so ignore
1. The first thing we have to do is to check how many of our libraries, if more than five, to see if it is the cause of the delay
2. Check to see if there are slow query statements, and if so, we'll optimize him.
3. See if buffer and cache are added to the front end
This article is from the "Wsyht blog" blog, make sure to keep this source http://wsyht2015.blog.51cto.com/9014030/1789148
13. The problem and solution of MySQL master-slave synchronization delay