We try it this way; Setup Vdisk swap as primary area then real disk swap space as #2 area (two swap disks) ... then set memory lower and lower until we start to see Vdisk swap some ... if we see swap move into the real disk area we then need to add a bit more memory until just Vdisk swap is used again ...
The swapon -s command shows the two swap spaces and their usage ... -----Original Message----- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Rich Smrcina Sent: Friday, April 11, 2008 8:18 AM To: [email protected] Subject: Re: FW: Could use some help tuning a POC One of the signs to knowing if Linux needs more memory is if Linux itself is paging. While your application is running use the vmstat command: 'vmstat 15' It will show some paging metrics (among others) in 15 second intervals. It will show up under the swap column. The first one line doesn't count because that's an average since the system came up. Pay attention to the rest though, while the app is running. If you see consistent high paging (swapping). You may want to bump up the size of the virtual machine. Do you have Vdisk set up for Linux swap? If you do that's great, because if Linux does need to swap it will do so very fast. Also, if the application demands more or runs out of memory, you'll *usually* get messages indicating such (from Linux or the application) or it may exhibit weird behavior. Bauer, Bobby (NIH/CIT) [E] wrote: > This was posted to both the VM and Linux listservs > > We finally got our POF running this morning and response is not great. > Since we know very little about this z/VM and zLinux I would appreciate > any hints. Performance Monitor shows some paging and swapping but from > our MVS point of view, its now bad. The VM guest only has 512M. I could > increase it but I keep hearing from Share and other seminars that > increasing memory is not usually the problem. There is 2G of memory > assigned to this lpar. > Some testers, at different geographical locations are getting good > results, others aren't so I'm looking into network problems also. > > Detailed data for user LKTEST2 > > > > > > Total CPU : ....% Storage def. : 512MB Page fault rate: > ..../s > > > > > Superv. CPU : ....% Resident <2GB: 129654 Page read rate : > ..../s > > > > > Emulat. CPU : ....% Resident >2GB: 0 Page write rate: > ..../s > > > > > VF total : ....% Proj. WSET : 131072 Pgs moved >2GB>: > ..../s > > > > > VF overhead : ....% Reserved pgs : 0 Main > XSTORE : > ..../s > > > > > VF emulation: ....% Locked pages : 13 XSTORE > main : > ..../s > > > > > VF load rate: ..../s XSTORE dedic.: 0MB XSTORE > DASD : > ..../s > > > > > I/O rate : ..../s XSTORE pages : 1314 SPOOL pg reads : > ..../s > > > > > DASD IO rate: ..../s DASD slots : 0 SPOOL pg writes: > ..../s > > > > > UR I/O rate : ..../s IUCV X-fer/s : ..../s MDC insert rate: > ..../s > > > > > Diag. X'98' : ..../s Share : 100 MDC I/O avoided: > ..../s > > > > > *BLOCKIO : ..../s Max. share : ... > > > > > > > > > > > #I/O active : 0 Active :100% PSW wait :100% I/O > act. : 0% > > > > > Stacked blk : .. Page wait : 0% CF wait : 0% > Eligible : 0% > > > > > Stat.: EME,DSC,PSWT I/O wait : 0% Sim. wait: 0% > Runnable : 0% > > > > > > > > > > Data Space Name Size Mode PgRd/s PgWr/s XRd/s XWr/s Migr/s > Steal/s > > > > > BASE 512MB Priv ... ... ... ... ... > ... > > > > > > > > > > Device activity and status: > > > > > 0009 3215 .0 000C 254R CL *, EOF NOH > NCNT > > > > Interval 07:45:21-07:46:23, on 2008/04/11 (CURRENT interval, select > interim <http://128.231.160.20:81/035986D8/3FB6/interim> or average > <http://128.231.160.20:81/035986D8/3FB6/average> data) > > > > > ______ . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.M_TOTAL> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.%25MUSED> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.H_TOTAL> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.%25HUSED> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.SHARED> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM./CAFREE> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.USED> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.S_TOTAL> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.%25SUSED> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.IN> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.OUT> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.CATES> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.MAJOR> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.MINOR> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.READ> . > <http://128.231.160.20:81/035986D8/3FB6/SO.LXMEM.WRITE> > > > > > <------------ Memory Allocation (MB) -------------> <------- > Swapping -------> <--- Pages/s ---> <-BlockIO-> > > > > > Linux <--- Main ---> <--- High ---> Buffers Cache <-Space > (MB)-> <-Pgs/sec-> Allo <-Faults--> <--kB/sec-> Nr of > > > > > Userid <http://128.231.160.20:81/035986D8/3FB6/HE.04.001> M_Total > <http://128.231.160.20:81/035986D8/3FB6/HE.04.011> %MUsed > <http://128.231.160.20:81/035986D8/3FB6/HE.04.019> H_Total > <http://128.231.160.20:81/035986D8/3FB6/HE.04.026> %HUsed > <http://128.231.160.20:81/035986D8/3FB6/HE.04.034> Shared > <http://128.231.160.20:81/035986D8/3FB6/HE.04.041> /CaFree > <http://128.231.160.20:81/035986D8/3FB6/HE.04.048> Used > <http://128.231.160.20:81/035986D8/3FB6/HE.04.057> S_Total > <http://128.231.160.20:81/035986D8/3FB6/HE.04.063> %SUsed > <http://128.231.160.20:81/035986D8/3FB6/HE.04.071> In > <http://128.231.160.20:81/035986D8/3FB6/HE.04.081> Out > <http://128.231.160.20:81/035986D8/3FB6/HE.04.086> cates > <http://128.231.160.20:81/035986D8/3FB6/HE.04.090> Major > <http://128.231.160.20:81/035986D8/3FB6/HE.04.096> Minor > <http://128.231.160.20:81/035986D8/3FB6/HE.04.102> Read > <http://128.231.160.20:81/035986D8/3FB6/HE.04.109> Write > <http://128.231.160.20:81/035986D8/3FB6/HE.04.114> Users > <http://128.231.160.20:81/035986D8/3FB6/HE.04.120> > > > > > >System< 499.4 89.2 .0 .0 .0 63.1 181.8 1344 > 2.7 .000 .000 74.28 .062 174.1 5.329 5.958 3 > > > > > LAMP1 <http://128.231.160.20:81/035986D8/3FB6/LXM.LAMP1> 499.4 > 93.6 .0 .0 .0 95.6 178.3 1024 .0 .000 .000 > 210.5 .000 503.1 .067 10.87 > > > > > LKTEST1 <http://128.231.160.20:81/035986D8/3FB6/LXM.LKTEST1> 499.4 > 75.7 .0 .0 .0 89.5 192.2 1504 .0 .000 .000 > .836 .000 .070 .000 2.648 > > > > > LKTEST2 <http://128.231.160.20:81/035986D8/3FB6/LXM.LKTEST2> 499.4 > 98.2 .0 .0 .0 4.3 174.9 1504 8.2 .000 .000 > 8.843 .182 12.94 15.60 4.231 > > Bobby Bauer > Center for Information Technology > National Institutes of Health > Bethesda, MD 20892-5628 > 301-594-7474 > > > ---------------------------------------------------------------------- > For LINUX-390 subscribe / signoff / archive access instructions, > send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit > http://www.marist.edu/htbin/wlvindex?LINUX-390 > -- Rich Smrcina VM Assist, Inc. Phone: 414-491-6001 Ans Service: 360-715-2467 rich.smrcina at vmassist.com http://www.linkedin.com/in/richsmrcina Catch the WAVV! http://www.wavv.org WAVV 2008 - Chattanooga - April 18-22, 2008 ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
