Sponsored Content
Operating Systems AIX Performance issue / tuning advice Post 302480507 by zaxxon on Wednesday 15th of December 2010 06:52:06 AM
Old 12-15-2010
I do not know SAS - I could see it is some Business Intelligence Software, which I guess can consist of different kinds of applications. More information would be interessting, if there is any type of database is involved, webservers etc.

Has anybody tuned this system already? There is a guide I just found which might be interessting for you:
http://www.sas.com/partners/director...stPractice.pdf

What does really not look good is:
  1. kthr/r Run Queue -- Did you try exporting AIXTHREAD_SCOPE=S? It looks to me, as if there are just far too few CPU resources to handle all ready-to-run threads. It would be really interessting to see, how the system behaves if you could add 3 more physical CPU units. Also it would be interessting, if your application benefits from partitioning your 3 physical CPUs to 6 virtual CPUs which would result with SMT turned on (which is the case in your system) to 12 logical CPUs instead of just 6.
  2. For scanning memory pages, the ratio is ok, example fr: 83624 sr: 126816 which is about 1:1,5. Though it is a lot of free pages the system demands and so with has to scan. Beside that it nearly always has about 80k pages on the free list. Could you please post the output of vmo -x minfree -x maxfree, thanks. Best add the output of vmo -x lru_file_repage too, just in case. Also the output of svmon -G might be interessting.
  3. What is definetly tunable, is the bad high counter for
    596244370 external pager filesystem I/Os blocked with no fsbuf. Setting the double of your current (check with ioo -x j2_dynamicBufferPreallocation) or if it is still 16 which is default, try 128 instead and check over time, if counter is still increasing by large number. You can also tune the ioo -x j2_nBufferPerPagerDevice to let's say 2048 and check if the counter increases rapidly still. It will take some time to monitor it. Check the man page for ioo, which of the 2 parameters can be changed dynamic and which needs a remounts of the filesystems.
  4. Currently your box is not paging in or out to Paging Space, but you have this: 731746 paging space I/Os blocked with no psbuf
    which shows, there has been such activity. I bet a lsps -a will show, that some of your Paging Space is used. If the output of ioo -x lru_file_repage turns out to be 0 already, then I would add some more memory to this box.

So far my 1st ideas for your box. Would be nice to hear any feedback if anything of it helped.

Last edited by zaxxon; 12-15-2010 at 07:57 AM.. Reason: typos, formatting
This User Gave Thanks to zaxxon For This Post:
 

9 More Discussions You Might Find Interesting

1. UNIX for Dummies Questions & Answers

Performance tuning.

can someone tell me a good site to go to in order to learn this. please do not recommen nay books because i dont have interest in that. if you know of any good sites with good straight forward explanation on how to split loads on machines that has excessive loading, please let me know Also,... (1 Reply)
Discussion started by: TRUEST
1 Replies

2. Filesystems, Disks and Memory

EXT3 Performance tuning

Hi all, long time ago I posted something, but now, it is needed again :( Currently, I am handling with a big NFS Server for more than 200 clients, this sever has to work with 256 NFSDs. Because of this huge amount of NFSDs, there are thousands of small write accesses down to the disk and... (3 Replies)
Discussion started by: malcom
3 Replies

3. UNIX for Dummies Questions & Answers

Performance Tuning

Hi to all, I'm interested in finding an introduction about Performance Tuning under Unix (or Linux); can somebody please point me in the right direction? Best regards (1 Reply)
Discussion started by: domyalex
1 Replies

4. Shell Programming and Scripting

Oracle-performance tuning

Sorry, This is out of scope of this group.But I require the clarification pretty urgently. My Oracle database is parallely enabled. Still,in a particular table queries do not work "parallely" always. How is this? (9 Replies)
Discussion started by: kthri
9 Replies

5. Shell Programming and Scripting

Performance Tuning

Hi All, In last one week, i have posted many questions in this portal. At last i am succeeded to make my 1st unix script. following are 2 points where my script is taking tooooo long. 1. Print the total number of records excluding header & footer. I have found that awk 'END{print NR -... (2 Replies)
Discussion started by: Amit.Sagpariya
2 Replies

6. UNIX for Advanced & Expert Users

Performance diagnosis & tuning

Hi, I am facing a strange issue. Application is deployed in a cluster with 2 Unix nodes (with same configuration). On one node the application is working fine but on another node we see this behavior I found using vmstat- when the server is not yet started everything is OK; when you start the... (3 Replies)
Discussion started by: ash.abrol
3 Replies

7. Solaris

Solaris Performance tuning

Dear all, I have a Local zone , where users feel that performance is not good. Is it wise to collect the inputs from the local zone rather than taking from the global zone. And also Can I tune from Global zone , so that it will reflect in local zone. Rgds rj (2 Replies)
Discussion started by: jegaraman
2 Replies

8. Tips and Tutorials

The Most Incomplete Guide to Performance Tuning

Overview: Introduction What Does Success Mean? What Does Performance Mean? Every Picture is Worth a Thousand Words Work Like a Physicist Work Like You Walk - One Step at a Time Learn to Know Your System Choose Your Weapons! Tools of the Trade 1 - vmstat A Little Theory Along the Way -... (1 Reply)
Discussion started by: bakunin
1 Replies

9. AIX

IBM AIX I/O Performance Tuning

I have a IBM Power9 server coupled with a NVMe StorWize V7000 GEN3 storage, doing some benchmarks and noticing that single thread I/O (80% Read / 20% Write, common OLTP I/O profile) seems slow. ./xdisk -R0 -r80 -b 8k -M 1 -f /usr1/testing -t60 -OD -V BS Proc AIO read% IO Flag IO/s ... (8 Replies)
Discussion started by: c3rb3rus
8 Replies
POSIX_MADVISE(3)					     Linux Programmer's Manual						  POSIX_MADVISE(3)

NAME
posix_madvise - give advice about patterns of memory usage SYNOPSIS
#include <sys/mman.h> int posix_madvise(void *addr, size_t len, int advice); Feature Test Macro Requirements for glibc (see feature_test_macros(7)): posix_madvise(): _POSIX_C_SOURCE >= 200112L DESCRIPTION
The posix_madvise() function allows an application to advise the system about its expected patterns of usage of memory in the address range starting at addr and continuing for len bytes. The system is free to use this advice in order to improve the performance of memory accesses (or to ignore the advice altogether), but calling posix_madvise() shall not affect the semantics of access to memory in the speci- fied range. The advice argument is one of the following: POSIX_MADV_NORMAL The application has no special advice regarding its memory usage patterns for the specified address range. This is the default behavior. POSIX_MADV_SEQUENTIAL The application expects to access the specified address range sequentially, running from lower addresses to higher addresses. Hence, pages in this region can be aggressively read ahead, and may be freed soon after they are accessed. POSIX_MADV_RANDOM The application expects to access the specified address range randomly. Thus, read ahead may be less useful than normally. POSIX_MADV_WILLNEED The application expects to access the specified address range in the near future. Thus, read ahead may be beneficial. POSIX_MADV_DONTNEED The application expects that it will not access the specified address range in the near future. RETURN VALUE
On success, posix_madvise() returns 0. On failure, it returns a positive error number. ERRORS
EINVAL addr is not a multiple of the system page size or len is negative. EINVAL advice is invalid. ENOMEM Addresses in the specified range are partially or completely outside the caller's address space. VERSIONS
Support for posix_madvise() first appeared in glibc version 2.2. CONFORMING TO
POSIX.1-2001. NOTES
POSIX.1 permits an implementation to generate an error if len is 0. On Linux, specifying len as 0 is permitted (as a successful no-op). In glibc, this function is implemented using madvise(2). However, since glibc 2.6, POSIX_MADV_DONTNEED is treated as a no-op, because the corresponding madvise(2) value, MADV_DONTNEED, has destructive semantics. SEE ALSO
madvise(2), posix_fadvise(2) COLOPHON
This page is part of release 4.15 of the Linux man-pages project. A description of the project, information about reporting bugs, and the latest version of this page, can be found at https://www.kernel.org/doc/man-pages/. Linux 2017-09-15 POSIX_MADVISE(3)
All times are GMT -4. The time now is 05:42 AM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy