2011/12/8 Piotr Kołaczkowski :
> Can someone tell me what is the use pattern of the CommitLog#add method? I
> mean, is it possible, that a single thread calls add many times, remembers
> the returned Future objects and *then* waits on all / some of them? Or is it
> always like: add, then wait (unti
W dniu 2011-12-08 08:40, Jonathan Ellis pisze:
2011/12/8 Piotr Kołaczkowski:
Right, this would be the best option to have an ability to write into
multiple log files, put on multiple disks. I'm not sure if it is part of
that ticket, though.
It's not. I don't think anyone needs more than 80MB/
2011/12/8 Piotr Kołaczkowski :
> Right, this would be the best option to have an ability to write into
> multiple log files, put on multiple disks. I'm not sure if it is part of
> that ticket, though.
It's not. I don't think anyone needs more than 80MB/s or so of
commitlog bandwidth for a while.
Right, this would be the best option to have an ability to write into
multiple log files, put on multiple disks. I'm not sure if it is part of
that ticket, though. Maybe we should split it into two things: parallel
serialization / CRC and parallel writes to multiple logfiles (as another
ticke
Another option is to have multiple threads reading from the queue and
writing to their own commit log files. If you have multiple commit log
directories with each having its own task writing to it, you can keep
the "only sequential writes" optimization. Multiple writers to one disk
only makes
