site_archiver@lists.apple.com Delivered-To: Macnetworkprog@lists.apple.com -josh On Sep 30, 2005, at 9:36 AM, Marc Krochmal wrote: On Sep 30, 2005, at 6:50 AM, Eric Dahlman wrote: Howdy, _______________________________________________ Do not post admin requests to the list. They will be ignored. Macnetworkprog mailing list (Macnetworkprog@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/macnetworkprog/site_archiver%40lists.... This email sent to site_archiver@lists.apple.com Something strange is happening on your network or machine. I talked briefly with Marc Krochmal. It sounds like the IP IDs are all the same. For each IP packet the stack sends, it picks a new IP ID. The stack is not generating these packets. You might check to see if you have any third party kernel extensions installed. I have sent the trace in a separate message. Just some extra information here. At present mDNSresponder is taking up about 50% of my CPU and is about 864 K in size. When I went home last night my powerbook did not show any of this traffic and when I came in to the office it also did not. I noticed it after about half an hour and switching users, sorry I cannot say exactly when it started or if fast user switching was a contributing factor. I also started a packet trace with Ethereal to watch http traffic during this time but alas it got too big to be manageable. I will describe those problems in my next installment as I think they have a different cause but still may be related ;-) The reason that mDNSResponder is using 50% CPU is not because it's sending these packets, it's because mDNSResponder is receiving and processing all these packets. Given you're having this problem and the "Slashdot" problem, I think this is happening somewhere in the kernel or in your Ethernet hardware. In the packet trace, every mDNS packet has the same IP identifier, meaning that this is the exact same packet being sent by the kernel over and over. Hopefully someone else on the list can provide more information about how to debug this. -Marc smime.p7s
participants (1)
-
Josh Graessley