[ From:	CRDGW2::CRDGW2::MRGATE::"SMTP::CRVAX.SRI.COM::RELAY-INFO-VAX" 31-OCT-1990 21:21:58.22  To:	MRGATE::"ARISIA::EVERHART" CC:	$ Subj:	Re: MPW parameters suggestion.  * Received:  by crdgw1.ge.com (5.57/GE 1.76)) 	 id AA17501; Wed, 31 Oct 90 21:06:05 EST Y Received: From UCBVAX.BERKELEY.EDU by CRVAX.SRI.COM with TCP; Wed, 31 OCT 90 15:34:52 PDT , Received: by ucbvax.Berkeley.EDU (5.63/1.42)* 	id AA22924; Wed, 31 Oct 90 15:18:42 -08009 Received: from USENET by ucbvax.Berkeley.EDU with netnews . 	for info-vax@kl.sri.com (info-vax@kl.sri.com); 	(contact usenet@ucbvax.Berkeley.EDU if you have questions)  Date: 31 Oct 90 23:05:47 GMTz From: usc!zaphod.mps.ohio-state.edu!uakari.primate.wisc.edu!dali.cs.montana.edu!milton!seymour@ucsd.edu  (Richard Seymour)/ Organization: University of Washington, Seattle ' Subject: Re: MPW parameters suggestion. + Message-Id: <10354@milton.u.washington.edu> . References: <11383E66FAFF003CC9@ewuvms.bitnet># Sender: info-vax-request@kl.sri.com  To: info-vax@kl.sri.com   o In article <11383E66FAFF003CC9@ewuvms.bitnet> DHELSING@EWUVMS.BITNET ("Dennis Helsing,  359-7080", 509) writes: F >I'm currently running VMS 5.2 on a 6410 and have had pretty good luckC >subscribing to the following rules of thumb for some MPW params...  >...some text deleted...7 >MPW_HILIMIT   = 15% of the total physical memory size.  >MPW_WAITLIMIT = MPW_HILIMIT$ >MPW_LOLIMIT   = half of MPW_HILIMIT >MPW_THRESH    = 16,384  > ? As he indicates in the deleted text, these are "rules of thumb" : We have a 3200-based  acquisition system with 8 megabytes  (so much for _THRESH :-)C but the way it manipulates data and global sections suggested (and  % performance indicated we were right)  E   MPW_LOLIMIT = 93   <-- autogen's default for "when to stop writing" E   MPW_HILIMT = 3000  <-- we want no writing-to-disk unless it's LARGE >   MPW_THRESH = 1000  <-- if there's a need, start writing hereM   MPW_WAITLIMIT = 6000  <--- don't freeze user processes unless REALLY tight! M   MPW_WRTCLUSTER = 96  <-- i think autogen's default -- compromise (2 tracks)   K the adjustment of WAITLIMIT is what most affected us -- we wanted processes J  to continue, even if the MPW list was getting pretty large (we have pulse3  loads which do not warrant freeze-to-flush action)     Your mileage WILL vary... 	 good luck  --dick  