Sponsored Content
Full Discussion: Scan Rates
Operating Systems AIX Scan Rates Post 302307784 by zaxxon on Thursday 16th of April 2009 10:29:18 AM
Old 04-16-2009
3k pages as scan rate per sec is no big value. I sometimes see bigger ones. Usually it's the ratio/relation between scanned pages and freed pages that matters. Optimal to ok should be ratio fr/sr of ~ 1:1 - 1:3 afaik; can be but must not.

From some IBM AIX Performance Management Guide, note the last line in bold letters:
Quote:
# fr

Number of pages that were freed per second by the page-replacement algorithm during the interval. As the VMM page-replacement routine scans the Page Frame Table (PFT), it uses criteria to select which pages are to be stolen to replenish the free list of available memory frames. The criteria include both kinds of pages, working (computational) and file (persistent) pages. Just because a page has been freed, it does not mean that any I/O has taken place. For example, if a persistent storage (file) page has not been modified, it will not be written back to the disk. If I/O is not necessary, minimal system resources are required to free a page.
# sr

Number of pages that were examined per second by the page-replacement algorithm during the interval. The VMM page-replacement code scans the PFT and steals pages until the number of frames on the free list is at least the maxfree value. The page-replacement code might have to scan many entries in the PFT before it can steal enough to satisfy the free list requirements. With stable, unfragmented memory, the scan rate and free rate might be nearly equal. On systems with multiple processes using many different pages, the pages are more volatile and disjoint. In this scenario, the scan rate might greatly exceed the free rate.

Memory is over committed when the ratio of fr to sr (fr:sr) is high.

An fr:sr ratio of 1:4 means that for every page freed, four pages had to be examined. It is difficult to determine a memory constraint based on this ratio alone, and what constitutes a high ratio is workload/application dependent.

For tuning in general and to get a first impression, you should look at the output from "vmstat 1 20" for example at busy times. Just looking at "sr" will show/mean nothing.
 

6 More Discussions You Might Find Interesting

1. UNIX for Dummies Questions & Answers

IP Name scan

Hi. how to search a range of IP:s for their registed IP names? Like nslookup or host for all IPs 130.xxx.xxx.1 to 130.xxx.xxx.254 //nicke (2 Replies)
Discussion started by: nicke30
2 Replies

2. Advertise with Us

Special Discount Rates for Job Posting In Effect

The UNIX and Linux Job Board is a service provided by the The UNIX and Linux Forums to help employers connect with UNIX and Linux professionals. All proceeds from this service go towards supporting the forums. Please PM Neo or email with any questions about this service. Promotion A: $99 ... (0 Replies)
Discussion started by: Neo
0 Replies

3. Shell Programming and Scripting

FTP run from shell script gives slow transfer rates

Hey everybody, this is my first post so be gentle. I have two Sun 5220's running Solaris 10 that are directly connected with a cross-over cable at Gig. One of these boxes is my production Oracle server which generates a 50GB dump file every evening at 10:50. The other Solaris is a devolopment... (8 Replies)
Discussion started by: Countificus
8 Replies

4. What is on Your Mind?

Where to find higher consulting rates?

Have any IT consultants here been on a project where you knew the bill rate was really high but you only got a tiny piece of it (like paid $60/hr and billed out around $200)? Does anyone know of a company that pays consultants well - like 70-80% or more of what they're getting? (5 Replies)
Discussion started by: apierce
5 Replies

5. Cybersecurity

Virus/Malware Uptake Rates

Hi. I'm trying to get information about the rates at which viruses and malware infect computers. Let's say that Mr. Dastardly Developer discovers an exploitable flaw, writes a virus to take advantage of it, and releases the virus. Assuming that a large-scale attack method is chosen (Facebook,... (2 Replies)
Discussion started by: treesloth
2 Replies

6. Hardware

BackBlaze article on HDD failure rates.

Very interesting. One for the pros here... Hard Drive Failure Rates: The Results from 68,813 Hard Drives (3 Replies)
Discussion started by: wisecracker
3 Replies
VMSTAT(1)						    BSD General Commands Manual 						 VMSTAT(1)

NAME
vmstat -- report virtual memory statistics SYNOPSIS
vmstat [-CefHiLlmstUvW] [-c count] [-h hashname] [-M core] [-N system] [-u histname] [-w wait] [disks] DESCRIPTION
vmstat reports certain kernel statistics kept about process, virtual memory, disk, trap, and CPU activity. The options are as follows: -C Report on kernel memory caches. Combine with the -m option to see information about memory pools that back the caches. -c count Repeat the display count times. The first display is for the time since a reboot and each subsequent report is for the time period since the last display. If no wait interval is specified, the default is 1 second. -e Report the values of system event counters. -f Report fork statistics. -H Report all hash table statistics. -h hashname Report hash table statistics for hashname. -i Report the values of system interrupt counters. -L List all the hashes supported for -h and -H. -l List the UVM histories being maintained by the kernel. -M core Extract values associated with the name list from the specified core instead of the default /dev/mem. -m Report on the usage of kernel dynamic memory listed first by size of allocation and then by type of usage, followed by a list of the kernel memory pools and their usage. -N system Extract the name list from the specified system instead of the default /netbsd. -s Display the contents of the uvmexp structure. This contains various paging event and memory status counters. -t Display the contents of the vmtotal structure. This includes information about processes and virtual memory. The process part shows the number of processes in the following states: ru on the run queue dw in disk I/O wait pw waiting for paging sl sleeping The virtual memory section shows: total-v Total virtual memory active-v Active virtual memory in use active-r Active real memory in use vm-sh Shared virtual memory avm-sh Active shared virtual memory rm-sh Shared real memory arm-sh Active shared real memory free Free memory All memory values are shown in number of pages. -U Dump all UVM histories. -u histname Dump the specified UVM history. -v Print more verbose information. When used with the -i, -e, or -m options prints out all counters, not just those with non-zero values. -W Print more verbose information about kernel memory pools. -w wait Pause wait seconds between each display. If no repeat count is specified, the default is infinity. By default, vmstat displays the following information: procs Information about the numbers of processes in various states. r in run queue b blocked for resources (i/o, paging, etc.) memory Information about the usage of virtual and real memory. Virtual pages (reported in units of 1024 bytes) are considered active if they belong to processes which are running or have run in the last 20 seconds. avm active virtual pages fre size of the free list page Information about page faults and paging activity. These are averaged every five seconds, and given in units per second. flt total page faults re page reclaims (simulating reference bits) pi pages paged in po pages paged out fr pages freed per second sr pages scanned by clock algorithm, per-second disks Disk transfers per second. Typically paging will be split across the available drives. The header of the field is the first charac- ter of the disk name and the unit number. If more than four disk drives are configured in the system, vmstat displays only the first four drives. To force vmstat to display specific drives, their names may be supplied on the command line. faults Trap/interrupt rate averages per second over last 5 seconds. in device interrupts per interval (including clock interrupts) sy system calls per interval cs CPU context switch rate (switches/interval) cpu Breakdown of percentage usage of CPU time. us user time for normal and low priority processes sy system time id CPU idle FILES
/netbsd default kernel namelist /dev/mem default memory file EXAMPLES
The command ``vmstat -w 5'' will print what the system is doing every five seconds; this is a good choice of printing interval since this is how often some of the statistics are sampled in the system. Others vary every second and running the output for a while will make it appar- ent which are recomputed every second. SEE ALSO
fstat(1), netstat(1), nfsstat(1), ps(1), systat(1), iostat(8), pstat(8) The sections starting with ``Interpreting system activity'' in Installing and Operating 4.3BSD. BUGS
The -c and -w options are only available with the default output. The -l, -U, and -u options are useful only if the system was compiled with support for UVM history. BSD
October 22, 2009 BSD
All times are GMT -4. The time now is 02:25 PM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy