On Fri, Feb 27, 2009 at 7:04 AM, Cui Shijun <rancp...@gmail.com> wrote: > hi, > I wonder if it is safe to assume that binlog can stay in master's > memory when replication happens
It's not safe to assume. It varies from system to system depending on operating system, filesystem, scheduler algorithm, amount of memory on the machine, and probably a bunch of other variables. The binlog is first and foremost a file, and any portion of it that happens to be cached in memory is an operating-system optimization that cannot be counted on. What does replication have to do with it? >. If not, when the binlog getts corruptted, will the slave's binlog also get >corrupted? Slave's binlog? The slave's binlog is not tied to the master's binlog. I am not sure you understand clearly how replication works. If the slave reads a corrupt binlog from the master's disk, then the slave's RELAY log will be corrupt too. > Is there way to make the slave's binlog survive even in master's disk > failure? Again I'm not sure what you are really asking. You might be looking for the Google synchronous replication patches. Do a web search for those and read up on what they offer. -- Baron Schwartz, Director of Consulting, Percona Inc. Our Blog: http://www.mysqlperformanceblog.com/ Our Services: http://www.percona.com/services.html -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/mysql?unsub=arch...@jab.org
