07-26-2013
This is an impression - not a solution:
You've got a series of issues, it looks like. You have several RT processes, those preempt everybody else. This is not necessarily always bad.
One problem appears to be context switching, often caused when nobody except high priority processes completes a quantum. Or there are loads of low priority processes that get the cpu frequently due to schedule policy and then get booted out.
During context switches, unless there is cpu affinity for the processes playing musical chairs, every context switch potentially involves cache latency. So you MAY get lots of cpu cache thrashing. Even with cpu affinity set you can get thrashing.
This is not so bad when a process gets a full 20ms (example value) quantum. It is murder when most processes get kicked out after 1ms. Because relatively much more cpu time is spent on loading cpu caches, not running the codestream.
IMO the large cpu wait queues speak to that.
Q: what exact OS are you running? and exact CPU's (NUMA?)
Q: what does this server do aside from run ruby, java and nfs?
Q: what is the output of uptime ( or who -r)
Q: can you run sar (set it up to gather data?). We can help
10 More Discussions You Might Find Interesting
1. UNIX for Advanced & Expert Users
Hi,
I am seeing very high kernel usage and very high load averages on my system (Although we are not loading much data to our database). Here is the output of top...does anyone know what i should be looking at?
Thanks,
Lorraine
last pid: 13144; load averages: 22.32, 19.81, 16.78 ... (4 Replies)
Discussion started by: lorrainenineill
4 Replies
2. Red Hat
Hi Buddies,
Thanx for reading my first post...
After googling a lot and searching so many forums I am feeling down a bit...
Please don't mind my ignorence, and my grammer ... :)
My server is running RHEL 2.6.9-5.EL. The cpu load is going higher than roof, almost 100 sometimes.
I am... (2 Replies)
Discussion started by: squid04
2 Replies
3. UNIX for Dummies Questions & Answers
Hello all, I have a question about load averages.
I've read the man pages for the uptime and w command for two or three different flavors of Unix (Red Hat, Tru64, Solaris). All of them agree that in the output of the 2 aforementioned commands, you are given the load average for the box, but... (3 Replies)
Discussion started by: Heathe_Kyle
3 Replies
4. Solaris
Hi All,
Please see to the prstat o/p of one of my sun box..
Total: 1 processes, 68 lwps, load averages: 531.00, 305.18, 144.77 Check the pstack ....
As i have read in all docs , people say a value of 5 is considered high CPU usage , i don't know then how we can even relate those... (3 Replies)
Discussion started by: mpics66
3 Replies
5. UNIX for Dummies Questions & Answers
How to determine what is causing high load average in a system? (3 Replies)
Discussion started by: proactiveaditya
3 Replies
6. AIX
Hi AIX Expert,
the fr (page freed/page replacement) and sr (pages scanned by page-replacement algorithm) values from the vmstat output (see below please) are very high. I usually see this high value during the oracle database backup. In addition, the page scan/page steal/ page faults values... (7 Replies)
Discussion started by: Beginer0705
7 Replies
7. UNIX for Dummies Questions & Answers
how load average is calculated and what exactly is it
difference between cpu% and load average (9 Replies)
Discussion started by: robo
9 Replies
8. Red Hat
i have a Intel Quad Core Xeon X3440 (4 x 2.53GHz, 8MB Cache, Hyper Threaded) with 16gig and 1tb harddrive with a 1gb port and my apache is causing my cpu to go up to 100% on all four cores heres my http.config
<IfModule prefork.c>
StartServers 10
MinSpareServers 10
MaxSpareServers 15... (4 Replies)
Discussion started by: awww
4 Replies
9. UNIX for Advanced & Expert Users
Hi all, hope you can help me. I'm getting high load average and can't find a reason for this, please share your inputs.
load average: 7.78, 7.50, 7.31
Tasks: 330 total, 1 running, 329 sleeping, 0 stopped, 0 zombie
Cpu0 : 7.0%us, 1.0%sy, 0.0%ni, 23.9%id, 0.0%wa, 38.9%hi,... (4 Replies)
Discussion started by: erick_tuk
4 Replies
10. UNIX for Advanced & Expert Users
With linux kernel 2.4.22-1.2199.nptlsmp (I know, it's very old) Sometimes Load average increases to big value (over 7) but my 4 vCPU are in
idle state (5% busy every cpu). My web procedure was gone down so I found out that process (with 4732 process id, see my following output)
was in... (4 Replies)
Discussion started by: zio_mangrovia
4 Replies
UPTIME(1) Linux User's Manual UPTIME(1)
NAME
uptime - Tell how long the system has been running.
SYNOPSIS
uptime
uptime [-V]
DESCRIPTION
uptime gives a one line display of the following information. The current time, how long the system has been running, how many users are
currently logged on, and the system load averages for the past 1, 5, and 15 minutes.
This is the same information contained in the header line displayed by w(1).
System load averages is the average number of processes that are either in a runnable or uninterruptable state. A process in a runnable
state is either using the CPU or waiting to use the CPU. A process in uninterruptable state is waiting for some I/O access, eg waiting for
disk. The averages are taken over the three time intervals. Load averages are not normalized for the number of CPUs in a system, so a
load average of 1 means a single CPU system is loaded all the time while on a 4 CPU system it means it was idle 75% of the time.
FILES
/var/run/utmp
information about who is currently logged on
/proc process information
AUTHORS
uptime was written by Larry Greenfield <greenfie@gauss.rutgers.edu> and Michael K. Johnson <johnsonm@sunsite.unc.edu>.
Please send bug reports to <albert@users.sf.net>
SEE ALSO
ps(1), top(1), utmp(5), w(1)
Cohesive Systems 26 Jan 1993 UPTIME(1)