Add support for --check-slave-lag for more than one server for pt-archiver

[Sveta Smirnova]

Currently pt-archiver supports option --check-slave-lag for single slave only. Add ability to specify more than one server this way.

[Sveta Smirnova (svetasmirnova)]

Version 2.2.14

[Michael Wang (xw73)]

I would request to implement the same solution as pt-online-schema-change with
--recursion-method dsn=D=<schame>,t=<table> to make it uniform across the tools.

Thanks.

[Frank Cizmich (frank-cizmich)]

Implemented ability to specify multiple slaves by being able to specify --check-slave-lag multiple times.
If --progress is defined, when pausing a report for each lagging slave is given every "--check-interval" seconds.

Note: The proposed --recursion-method approach was more involved to implement, but might be considered later on.

[Frank Cizmich (frank-cizmich)]

command line example for two slaves:

pt-archiver --no-delete --where '1=1' --source h=oltp_server,D=test,t=tbl --dest h=olap_server --check-slave-lag h=slave1 --check-slave-lag h=slave2 --limit 1000 --commit-each

[Dave Gregg (dgregg)]

Slave Lag now works well for multiple Slaves.....but

No matter what - row-by-row or BULK
you have records left and not copied over........look at bug Bug #1096274
I put comments in there as that bug is more specific I believe.

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PT-683