Date: Tue, 29 Mar 2005 11:31:30 -0500 From: Frank Laszlo <laszlof@tvog.net> To: Ken Smith <kensmith@cse.Buffalo.EDU> Cc: hubs@freebsd.org Subject: Re: CVSup port upgrade Message-ID: <424982E2.4050802@tvog.net> In-Reply-To: <1112109734.19982.46.camel@opus.cse.buffalo.edu> References: <XFMail.20050325134703.jdp@polstra.com> <1112109734.19982.46.camel@opus.cse.buffalo.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
The load for cvsup13.us is pretty low. I am averaging only about 5 or so users. I can probably take on another top level host on the same system. Its a Dual 1.8Ghz Xeon with 2G of RAM, and Ultra160 SCSI disks w/ RAID5. So its plenty beefy enough to handle it. Let me know if you are interested. __________________________________________________ Frank Laszlo System Administrator The VonOstin Group Email: laszlof@tvog.net WWW: http://www.vonostingroup.com Mobile: 248-863-7584 Ken Smith wrote: >On Fri, 2005-03-25 at 13:47 -0800, John Polstra wrote: > > > >>The fix for the bug is in the cvsupd server. Lots of people >>would appreciate it if you'd upgrade your cvsup installation to >>"cvsup-16.1h_2" (or the cvsup-without-gui port with the same >>revision). >> >> > >The three machines I more or less watch over: > > cvsup{5,8,18}.freebsd.org/cvsup{9,18}.us.freebsd.org > cvsup9.freebsd.org > cvsup10.{us.}freebsd.org > >should be all set. I'll do cvsup-master some time soon, given its >clients this bug shouldn't be biting any of them (mostly just mirror >sites). > >That first one on the list could still use a little help... It's what I >point existing names at when a site disappears on us. Of the names it's >currently supporting we could use: > > 1) Two new sites in the USA to take over cvsup{9,18}.us.freebsd.org > 2) Two sites anywhere in the world to take over > cvsup{8,18}.freebsd.org. > >For (2) we would be looking for already existing sites that have a >large-ish server set up for it but you feel it is currently severely >under-utilized. We would like the servers in the TLD to be established >sites with a good track record, good network connectivity, and capable >of handling a fairly significant load. Based on observing the three >machines I watch over it's best if the machine have 1Gb to 2Gb of RAM >and, if possible, SCSI disks (SCSI disks is less important if it's got >2Gb memory). With that the machine can typically handle 15 to 25 >simultaneous clients and still remain "responsive" (meaning it doesn't >take forever for each client to finish its updates). With less than 1Gb >RAM the servers tend to become *severely* I/O bound. > >If you'd like to volunteer for any of the above let me know. :-) > > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?424982E2.4050802>