On 25-Jun-2004 Robert Watson wrote: > > It might be interesting to look at the following: > > - Use sockstat/netstat to identify the fetch socket, and see whether > either the send of the receive queue contains some amount of > persistent, un-processed data. This would suggest if the > application was stalling and not reading, or that TCP was stalling > and not sending. > > - Use DDB to generate a stack trace of the fetch process in-kernel, > perhaps a few to see whether it's stuck in one place, and if so, > where. This would tell us what it's stuck doing. > > - Use ktrace to generate a trace of fetch and "see what it's doing" > when it appears to hang -- is it looping waiting for I/O, just > blocked in kernel, etc. Good suggestions all. Thanks, I'll try those. Incidentally, it does seem to be working better this morning. Perhaps it *was* just an ISP glitch. -- Conrad J. Sabatier <conrads_at_cox.net> -- "In Unix veritas"Received on Fri Jun 25 2004 - 12:28:39 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:58 UTC