> From: tech-boun...@lists.lopsa.org [mailto:tech-boun...@lists.lopsa.org] > On Behalf Of Edward Ned Harvey (lopser) > > OMG, holy crap. I don't care how big your processes are, or how much > memory you have in your system, you should never be swapping active > memory. You can always solve this problem by either adding more memory, > or using a memory mapped file in your process.
I forgot swap is religion. Hey wait - I'm going to print & sell t-shirts at the next meeting. ;-) Anyway, if you're not swapping much, maybe it's not worth while for the developers to figure out memory mapping. You have to ask yourself: Which wastes more of the developers' time? The time waiting for thrashing, or the time to figure out memory mapping? Or, if the application isn't in your control, maybe the application licensing cost and engineer time to use the tool don't outweigh the cost of a new high mem server. You have to ask yourself: Which wastes more money? Engineers and sales people waiting for the trashing to finish, plus whatever is the cost for the tool you're using? Or the cost of buying a new server with more memory? Is the time to completion of this project on the critical path for marketing or product delivery deadlines? Which is more significant? So I concede - I spoke too absolutely. My actual opinion is pretty extreme, but not absolute. One really obvious situation where it's not worth while to figure out memory mapping or pay for hardware upgrades would be the unpaid research project of some intern using a free tool... And of course, there's a whole gray area in between the extremes. I happen to support a lot of high mem high performance jobs, including EDA tools, with sims that will consume 100G+ memory... I haven't yet seen the situation where we deemed the cost of the hardware upgrade to be the most significant factor. _______________________________________________ Tech mailing list Tech@lists.lopsa.org https://lists.lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
