My transaction logs became corrupted and I had to skip 50 insert records from master. I've exported the data and am planning to update the slaves manually, with the slave stopped. The auto-increment value on the slave stayed in sync with master to my surprise so I think I can just re-populate the missing rows. Are there any potential issues with this plan?
So for example on master I run:
SELECT count(*) FROM `users` WHERE userid between 20 and 69;
and I get back 50. When I run it on slave I get "Empty set". So I did:
SELECT * FROM `users` WHERE userid between 20 and 69;
Then constructed the resulting CSV from PHPMyAdmin into an insert query:
insert ignore into users (userid, name, passwordhash, active) values (....,...,...,...)
Prior to insert on the slaves I will run:
and then after
I can't think of issues with this may cause but I don't have a testing replication environment so can't test this thoroughly.