FAQ

EN / [JP](FAQ JP)

index server seems to be Single Point Of Failure

It's true and considered as future tasks, but actally down of index server is not so critical. All required data is stored at /path/to/data/dir/flare.xml, so if index server is down, please calm down and restart server or process and everything is ok.

any actual performance ?

Flare is running at GREE (one of the major SNS services in Japan) w/ 6 masters and 6 slaves (over 2,000M keys), and 500-1,000 qps (load average is nearly 0.00...).