RedHat AS 2.1 Question - Memory Usage

Post #170,913 by gdaustin 8/25/04 11:56:58 AM Reply	RedHat AS 2.1 Question - Memory Usage We just got 2 new IBM x445 2way systems with 4gig of memory at work. Our applications team just got the boxes yesterday, and we ran top before we started installing our applications. We ran top and found 3.4 - 3.6 gig of our real memory in use, before we even load applications! Our current development system, an x335 box, has 2 gig of memory and we don't even use 1 gig with applications loaded, so something is wrong! However, when we sort the top by memory, the largest RSS is about 12-13meg. And there aren't enough to add up to 4gig, not even close. (And some are multithreaded apps, which report 13gig for each thread.) Is there some other way to look at the memory to see if top is even correct? I remember commands in Solaris (but I can't remember the EXACT command) that gave me the pages per application, and total usage at the top. I thought maybe it was a vmstat with some options, but I haven't got it to work on the RH AS 2.1. Ideas? Glen
Post #170,920 by Yendor 8/25/04 1:06:31 PM Reply	"top" reporting incorrect Especially on multi-CPU boxes. At least, that's the reports that I've gotten from multiple people at my office where we have a few 4xCPU Dells loaded with RHES (not RHAS). I haven't bothered to figure out what the correct command would be for that kind of reporting, mainly because I haven't needed to. But don't believe everything that `top` tells you on your multi-CPU box. -YendorMike "They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." - Benjamin Franklin, 1759 Historical Review of Pennsylvania
Post #171,005 by broomberg 8/25/04 10:34:04 PM Reply	Bullshit Go grab my example below and send it to them.
Post #171,002 by gdaustin 8/25/04 10:10:16 PM 8/25/04 10:19:03 PM Reply	Possible Answer I was told by a "systems expert" that the cache was probably using most of the memory. Apparently in times of low cpu utilization, but higher i/o, the system will allocate memory for cache. As long as there aren't a lot of applications demanding memory, it will use most of the memory for cache. I don't remember the cache value before we reset the system, but after the reset, the memory stayed in 200-400 meg range. As the memory rose, the cache rose also. This particular expert said that the behavior is common in Linux RH AS systems and in Solaris. Apparently, had we not panicked on this and started loading up the system with applications, the cache would have washed out. Glen Edited by gdaustin Aug. 25, 2004, 10:19:03 PM EDT Expand All History
Post #171,004 by broomberg 8/25/04 10:33:07 PM Reply	Yup, relax Freshly installed box: \n 22:30:08 up 5:03, 2 users, load average: 0.00, 0.00, 0.00\n43 processes: 42 sleeping, 1 running, 0 zombie, 0 stopped\nCPU states: cpu user nice system irq softirq iowait idle\n total 0.1% 0.0% 0.0% 0.0% 0.0% 0.2% 99.7%\n cpu00 0.2% 0.0% 0.0% 0.0% 0.0% 0.2% 99.6%\n cpu01 0.0% 0.0% 0.0% 0.0% 0.0% 0.2% 99.8%\nMem: 1028520k av, 317272k used, 711248k free, 0k shrd, 46052k buff\n 147508k actv, 123548k in_d, 276k in_c\nSwap: 2044072k av, 0k used, 2044072k free 207588k cached\n Note the free amount with nothing running. So I then create a large file and DDed it to /dev/null: \n 22:30:08 up 5:03, 2 users, load average: 0.00, 0.00, 0.00\n43 processes: 42 sleeping, 1 running, 0 zombie, 0 stopped\n 22:33:20 up 5:06, 2 users, load average: 1.31, 0.58, 0.21\n44 processes: 43 sleeping, 1 running, 0 zombie, 0 stopped\nCPU states: cpu user nice system irq softirq iowait idle\n total 5.0% 0.0% 7.5% 0.0% 0.0% 85.0% 2.5%\n cpu00 5.0% 0.0% 10.0% 0.0% 0.0% 85.0% 0.0%\n cpu01 5.0% 0.0% 5.0% 0.0% 0.0% 85.0% 5.0%\nMem: 1028520k av, 1012520k used, 16000k free, 0k shrd, 48304k buff\n 188948k actv, 728000k in_d, 14256k in_c\nSwap: 2044072k av, 0k used, 2044072k free 881060k cached\n Memory all gone! Don't worry, it is used by the cache and will be returned to the system as needed.
Post #171,029 by folkert 8/26/04 10:25:36 AM Reply	Sometimes... too much cache and not enough buffers... can cause serious delays. But not until it is massively using stuff. -- [link\|mailto:greg@gregfolkert.net\|greg], [link\|http://www.iwethey.org/ed_curry\|REMEMBER ED CURRY!] @ *i_wethey* No matter how much Microsoft supporters whine about how Linux and other operating systems have just as many bugs as their operating systems do, the bottom line is that the serious, gut-wrenching problems happen on Windows, not on Linux, not on Mac OS. -- [link\|http://www.eweek.com/article2/0,1759,1622086,00.asp\|source] Here is an example: [link\|http://www.greymagic.com/security/advisories/gm001-ie/\|Executing arbitrary commands without Active Scripting or ActiveX when using Windows]

Welcome to IWETHEY!