Hi Simon, See http://www.datastax.com/dev/blog/repair-in-cassandra for a good explanation. Basically repair is not a good word because it suggests it is fixing something that is broken. This is not the case. It ensures the consistency across your cluster. Not running repairs is a VERY BAD THING. Key things to check are how long do your repairs take and do they affect the operation of your cluster. Decide how often and when to run them based on this. The general rule of thumb is to run them within the gc_grace_seconds. Another good article is *Cassandra’s “Repair” Should Be Called “Required Maintenance”* which is available at https://lostechies.com/ryansvihla/2015/09/25/cassandras-repair-should-be-called-required-maintenance/
Best wishes, Ben On 18 November 2016 at 09:41, wxn...@zjqunshuo.com <wxn...@zjqunshuo.com> wrote: > Hi All, > I'm new to Cassandra and from the mail chain seemed repair is an important > thing to do. To avoid trouble in putting Cassandra in production > environment, I have some questions. > 1. What exactly does Cassandra repair do? > 2. I saw someone do repair in scheduled time, daily or weekly. Is it a > must to do this? > 3. In which cases do I need to do repair? > > Cheers, > -Simon Wu >
