Re: lsof befuddled on 10.4.6?
site_archiver@lists.apple.com Delivered-To: Darwin-dev@lists.apple.com Charles Thanks, Vincent On Apr 24, 2006, at 5:01 PM, Charles Bailey wrote: At no time does it return more than a few lines of data. Thanks for any help you can offer. _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.appl... I call this a bug that you should report at http:// bugreporter.apple.com/ I've been seeing a lot of unexplained disk activity (based on my observation of the drives' LEDs) and CPU system time on one of the 10.4.6 systems here -- dual G5 with 3.5 GB RAM, 2 external FW drives, and a pair of internal SATAs bound as a software RAID1. I first tried top, which didn't give any clear leads, then thought lsof might give me a decent start at figuring out whether this was mdimport or NAV or some other process. However, invoking /usr/sbin/ lsof as root produces inconsistent results, with one of the following occurring: - No processes listed; no error messages - No processes listed; error message "can't read process table" - A small but variable number (2-8 or so) of files listed, derived from 1-3 processes; no error message - A small but variable number (2-8 or so) of files listed, derived from 1-3 processes; interspersed "can't read fileglob for . . ." errors As noted, the system's running 10.4.6 (Darwin xxxx 8.6.0 Darwin Kernel Version 8.6.0: Tue Mar 7 16:58:48 PST 2006; root:xnu-792.6.70.obj~1/RELEASE_PPC Power Macintosh powerpc), with anywhere from half to >90% of the RAM in use, but a pagein rate of 0-50/sec. Is this expected behavior for lsof, and if so is there a workaround? I hesitate to call it a bug since I'm not that deeply familiar with OS X internals. This email sent to site_archiver@lists.apple.com
participants (1)
-
Vincent Lubet