No kidding, 50 was a small test. I do plan to utilize the bittorrent method soon, but haven't yet. I wish I could go into more detail, just like I wish I could send code patches :(. I'm working on getting through the red tape here to allow me to submit patches, but I'm limited to just subtle hinting for now...
I have tested the fix and am using it currently on my system. PERL's rename is a direct call into the system rename, which is atomic. I think this is just the "right way" (tm). Thanks, Matt On Wednesday 02 August 2006 10:01, Andrea Righi wrote: > wow! are you monitoring 50 clients? :-) have you tested the bittorrent > transport? > > Anyway the use of a temporary file seems reasonable, but have you tested > it or it's an hypothesis? do you have already a patch? > > Thanks, > -Andrea > > Matt Jamison wrote: > > I'd like to suggest the following change in si_monitor, after some heavy duty > > pounding... > > > > Rather than opening (truncating) the xml databasefile directly on each > > update, how about writing to a tmp file and then renaming the tmp file over > > the top of the actual database? With just 50 simultaneous clients, the > > truncation is quite noticeable. PERL's rename should be used to make the > > update atomic... > > > > Thanks, > > > > Matt > ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Sisuite-devel mailing list Sisuite-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sisuite-devel
