Date: Mon, 13 Nov 2006 09:59:48 +0200 From: Achilleas Mantzios <achill@matrix.gatewaynet.com> To: freebsd-java@freebsd.org Subject: Re: Performance of Java on FBSD vs. others... Message-ID: <200611130959.49073.achill@matrix.gatewaynet.com> In-Reply-To: <20061110223236.GD72658@ace.b020.ceid.upatras.gr> References: <20061110203714.GA89006@ace.b020.ceid.upatras.gr> <4554F4C6.3090802@infracaninophile.co.uk> <20061110223236.GD72658@ace.b020.ceid.upatras.gr>
next in thread | previous in thread | raw e-mail | index | archive | help
=CE=A3=CF=84=CE=B9=CF=82 =CE=A3=CE=AC=CE=B2=CE=B2=CE=B1=CF=84=CE=BF 11 =CE= =9D=CE=BF=CE=AD=CE=BC=CE=B2=CF=81=CE=B9=CE=BF=CF=82 2006 00:32, =CE=BF/=CE= =B7 Nikos Ntarmos =CE=AD=CE=B3=CF=81=CE=B1=CF=88=CE=B5: > Hi Matthew. > > On Fri, Nov 10, 2006 at 09:53:10PM +0000, Matthew Seaman wrote: > > Is this the same issue with syscalls as affects MySQL? As in > > unconstrained calling of gettimeofday() or similar because such > > "syscalls are free"? Which is pretty much true in Linux, at the > > cost of not returning the time particularly accurately. > > > > Look at http://wikitest.freebsd.org/MySQL -- especially the point > > about "frequent queries of the system time". Try tweaking the > > 'kern.timecounter.choice' sysctl and see if that makes much > > difference. > > That was my first thought too. I've tried with ACPI-fast, TSC, and i8254 > and saw no noticeable difference (btw k.t.hardware is the one to be set > to whatever value and k.t.choice the one to be queried for available > time counters)... > =CE=9D=CE=AF=CE=BA=CE=BF, =CE=BA=CE=B1=CE=BB=CE=B7=CE=BC=CE=AD=CF=81=CE=B1, just a question: Did you benchmark you whole system (apart from java) using some=20 well known benchmark? =2Dserver 1.4.2 (even linux-jdk) on my workstation some times runs faster t= han=20 our Linux dual xeon 3.2GHZ server. You can even test by writing a simple cpu-intensive C program by yourself. Also i dont remember answering the /etc/malloc.conf question. Linking this to "aj" could make a *huge* difference (if you are on current). Also trying removing any debugging from your kernel. man 7 tuning is a good way to start. If your system proves to be *generally* slower than the rest 2, then we got a FreeBSD issue here. else it is a java-related matter. Anyway in the meantime i wish you to solve the problem and stick all the way with FreeBSD! > \n\n > _______________________________________________ > freebsd-java@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-java > To unsubscribe, send any mail to "freebsd-java-unsubscribe@freebsd.org" =2D-=20 Achilleas Mantzios
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200611130959.49073.achill>