Shulou(Shulou.com)06/01 Report--

How mha works

1. Inspection phase

. Maser goway detected

. Multi-route detection

. Repeat detection 4 times (this number is fixed and cannot be modified)

. Check whether the ssh is normal

. Check the configuration (mainly detect which machines are in the cluster, which can be accessed normally, and which are already dead,binlog

Candidate master,relay_log_purge,read_only)

. Check the master status again

two。 Shut down the service phase

To prevent data anomalies, shut down the server phase. You can call the master_ip_failover_scripts script to shut down the server

3. Recovery phase

(1) .master recovery phase

. Find out the latest slave and the oldest slave.

. Copy the unsent binlog and save it on manager (T1).

. Use the latest slave to restore other slave so that all slave are at the same point in time

. Select a new master

. Compare T1 with master to generate differential binlog, redo it, and save binlog and pos (T2).

At this point, you can call master_ip_failover_scripts to open the service to the outside world.

(2) recovery stage

Compare T1 and slave in parallel to generate differential binlog and redo

(3). Modification stage of information

(1)。 Execute reset slave all on newmaster

(2)。 Perform change master binlog on other slave and pos on T2

(3)。 Modify mha configuration file

4. Inspection phase (reporting phase)

You can call the report_scripts script to check whether the new cluster is normal.

The details of the log are as follows:

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.