 |
 |
Your ticket #303 seems to be a pretty valid explanation for that. You should have a look at how much data you get per event triggered. Each event results with quite a lot of locking in the client code (both IOEvents and client proper), hence the futex calls shown by strace. Note that locking/unlocking an uncongested mutex does not result with a syscall in Linux. I think we would not see the problem if watermarking worked properly. In principle there is also no reason why you cannot have multiple poller threads as long as all the parallel stream associated with one session are assigned the same thread. --- Reply to this email directly or view it on GitHub: https://github.com/xrootd/xrootd/issues/20#issuecomment-158890653 ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the XROOTD-DEV list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1