Date: Tue, 24 Nov 2009 14:15:21 -0500 From: stan <stanb@panix.com> To: Free BSD Questions list <freebsd-questions@freebsd.org> Subject: This is to good to be true Message-ID: <20091124191521.GA9416@teddy.fas.com>
next in thread | raw e-mail | index | archive | help
I have about a dozen FreeBSD boxes serving a dedicated function. Among other things these machines run a local copy of Firefox, which is updated rapidly by a CGI script. These machines were at 6.2 STABLE, and Firefox 1.x. I am upgrading them to 7.2 STABLE, and Firefox 3. Historically memory leaks in Firefox cause us to have to write a "watchdog" script that killed it an restarted it about once a day, based upon it's active memory set. One of the new machines has been up for several days without having to do this. This is not totally unexpected,as I thought that this had bee improved, if not fixed in newer versions of Firefox. However, I have been keeping a close look on memory utilization using cricket (which acquires it's data using SNMP) and I am blown away by what I am seeing historically ucd_sys free ram has hovered around zero on these systems, Now it grows over time! These machines do have a stable set of applications, and I can rationally see how a well designed kernel might be aggressively freeing RAM given this. If so huge congratulations are due the developers. Dose this seem to be a reasonable belief? -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20091124191521.GA9416>