NYCPHP Meetup

[Adam Fields]

On Wed, Dec 28, 2005 at 04:02:32PM -0500, David Mintz wrote:
> Oops, sorry, too late. I restored replication yesterday and don't
> remember.  Other than "seconds behind master" was NULL and some other
> value up at the top of the output was other than "waiting for master to
> send event," and I did a couple informal tests -- change some data on
> master, see if update propagates to slave -- and they failed.

Okay, well, troubleshooting replication is usually pretty
easy. There's not much to it. The master generates binlog files when
queries are run, and the slave pulls these over and runs them.

If there's a problem, it's almost certainly one of the following:

1) The slave is trying to log in to the master with a login that
   either doesn't exist, has the wrong password (or hash scheme), or
   has insufficient permissions.

2) There's some network problem that keeps the slave from seeing the
   master, or the ip address/port of the master is wrong.

3) The binlog files aren't being read, written, or run properly.

4) The binglog files can't be run because the slave state is out of
   sync with the state of the master when then binlog files were
   generated or the slave is using the wrong binlog file/position.

That's basically it, and it's usually relatively easy to narrow down
where the problem is.

-- 
				- Adam

** Expert Technical Project and Business Management
**** System Performance Analysis and Architecture
****** [ http://www.everylastounce.com ]

[ http://www.aquick.org/blog ] ............ Blog
[ http://www.adamfields.com/resume.html ].. Experience
[ http://www.flickr.com/photos/fields ] ... Photos
[ http://www.aquicki.com/wiki ].............Wiki
[ http://del.icio.us/fields ] ............. Links