Hey, John. Who are you talking to? Gee the lists are quiet.
Andrew 8) -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John Tolmachoff (Lists) Sent: Thursday, February 10, 2005 9:46 AM To: [email protected] Subject: RE: [sniffer] Lists Ping? Your ping was not received. You must have done something wrong. No one is here. No one is home. :\ John Tolmachoff Engineer/Consultant/Owner eServices For You > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On > Behalf Of Marc Catuogno > Sent: Thursday, February 10, 2005 9:35 AM > To: [email protected] > Subject: [sniffer] Lists Ping? > > Is it just me or are all the lists (Imail, Declude V and JM and this > one > offline??) > > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] > On Behalf Of Pete McNeil > Sent: Tuesday, February 08, 2005 5:18 PM > To: Bill Green dfn Systems > Subject: Re: [sniffer] ERROR message in snifferp Command Prompt window > > On Tuesday, February 8, 2005, 3:20:25 PM, Bill wrote: > > BGdS> I have started seeing this line repeated in the persistent > BGdS> sniffer > command > BGdS> window. > > BGdS> ERROR_LOGFILE: Bad Lock During Logging > BGdS> c:\imail\declude\sniffer\"mycode".log > > BGdS> It looks like the error has been happening once a day for about > BGdS> a > week. > BGdS> Other than the message all seems to be working well. Where > BGdS> should I > look for > BGdS> the cause? > > The first clue I can see is that it happens once per day... Chances > are there is a scheduled process interfering with the log file, the > storage system in general (perhaps some backups or other IO intensive > operation). > > Locking is a very lightweight mechanism in SNF because most operations > are synchronized and sequential. If you are only seeing one of these > per day then there is no cause to worry - but do keep an eye on it so > that it doesn't get worse without you knowing it. > > A bad lock is probably a stale lock file --- The protocol would be to > simply ignore the lock after waiting the appropriate amount of time. > > In theory, no lock should be required to write to the log file because > it is opened in "append" mode. Unfortunately on Win32 based systems > this doesn't mean what it should. That is, write operations are not > 'atomic' --- so if more than one process tries to append to the log > file at once the result is unpredictable corruption. > > The locking mechanism we're using here (creating a lock semaphore > file) is only intended to synchronize access to the file since Win32 > doesn't. The fact that one process will wait - even if the lock fails > - usually accomplishes this task. If the process were to fail and two > processes wrote (append) to the log file at once then it is possible, > but not certain, that log corruption would occur -- which is not > strictly vital for the odd record here and there. > > Hope this helps, > > _M > > > > This E-Mail came from the Message Sniffer mailing list. For > information and > (un)subscription instructions go to > http://www.sortmonster.com/MessageSniffer/Help/Help.html > --- > [This E-mail scanned for viruses by Declude Virus] > > > > --- > [This E-mail scanned for viruses by Declude Virus] > > > This E-Mail came from the Message Sniffer mailing list. For > information and > (un)subscription instructions go to > http://www.sortmonster.com/MessageSniffer/Help/Help.html This E-Mail came from the Message Sniffer mailing list. For information and (un)subscription instructions go to http://www.sortmonster.com/MessageSniffer/Help/Help.html This E-Mail came from the Message Sniffer mailing list. For information and (un)subscription instructions go to http://www.sortmonster.com/MessageSniffer/Help/Help.html
