Can someone describe what the difference is between the select/accept/mutex behavior of .28 and .29/.30 on daedalus?
Also, do we have any data on syscalls per request during the spikes? Based on Greg's vmstat data showing a spike in syscalls/second concurrent with the spike in load, I'd expect that either: * The number of requests/second increased, or * The number of syscalls/request increased I'd originally assumed it was the latter (e.g., because some request type started doing more syscalls in .29 than in .28), but I can't yet identify evidence of this in the data. --Brian Sander Temme wrote: >>the point is I don't see the high load average. >> > >If we're actually looking at a problem with select()/accept() and the >presence or absence of a mutex, I would not be surprised if Solaris and BSD >react differently. See my December paper > ><http://research.covalent.net/projects/osdl1.html> > >For a look at the differences between Solaris and Linux. Now, Linux ain't >BSD, so it's possible that it does something else yet again. > >S. >
