> Yeah, it looks like the rados aio interface is racy. You should be > able to first set aio completion and only then to do the read/write > and not the other way around as it is now. I've just pushed a fix to > the unstable branch that changes the interface a bit. Now you first > need to do a rados_aio_create_completion() first and use it in the > rados_aio_read()/write() call. Note that I haven't tested it yet, so > let me know if there's any problem. > > Thanks, > Yehuda
Thank you - the race condition seems to be fixed. I'm still experiencing some cases where the callback function is being called a few hundred times, but I've to dig into this a little bit, before I can tell about it. Christian -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
