[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Thu, Mar 27, 2014 at 12:41:50AM +0000, bad apple wrote: > Thanks for yet another more balanced and equanimous reply than my > slightly tetchy post warranted... you just never get your rage face on > in public forums do you? Whilst I normally manage to not fly off the > handle I can't help being a bit barbed sometimes. Stupid internet is > slowly eroding my manners it seems. Heh. I don't get angry easily, and when I do I can't form a coherent sentence, so in practice I end up not showing any rage on the Internet. > Well it seems at this point everyone - especially you, which helps - is > firmly pointing their finger at your code as the problem, which is fair > enough. Unless it's sensitive in some way, why not do the usual which > these days is posting it on github? I can't publish the code at the moment for a number of complicated reasons, but even if I could, we're talking about 10k+ lines of code, divided over dozens of libraries and scripts, using a number of other programs. It would take someone a lot of time to understand what's going on. (Which, obviously, is another problem.) Also, much as I know I'm neither the most talented programmer, nor the most brilliant sysadmin in the world, I'm not an asbolute beginner either. And I think it'd be unlikely for someone to spot some issues easily. > Have you had a *really* good bash at the problem with iostat and > vmstat? I've spent a lot of time looking at things and have made many changes based on what these told me. They've made some differences, but never big enough - and in some cases the improvements didn't last. Having said that, I got so fed up that I decided to spend an hour or two this morning seeing if I could improve performance. Guess what, I found something (an unreasonably big rsync job) that I could change and that has improved things quite significantly. (iostat showed me that IO load was extremely high during the first half of every minute, so I concluded it must be a cron job. After some time and some serious debugging, I found the culprit.) I would be overjoyed, if it wasn't for the fact that a) the rsync was added to the system long after the IO problems started (it was added as part of something that was supposed to reduce the problems) and b) I've seen significant improvement in the past that in the end didn't last. But we'll see. I only needed a somewhat temporary solution anyway. This might do. Thanks! Martijn. -- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/listfaq