Contents 1 To add/remove CPU on a ruining zone (dedicated-cpu) 1.1 First get the pset name 1.2 Before the change just 4 cpu are configured 1.3 To add/modify the numbers of CPU to 8 1.4 After the change 2 To add/modify CPU on a ruining zone (capped-cpu) 2.1 Before the change 2.2 To make the …
Solaris Dynamic Zone Resource Capping With Examples Read More »
Tuning java and GC The link below describes in great detail how to troubleshot, tune Java Garbage Collector(GC). Garbage First Garbage Collector Tuning
A controlled CPU HOG load test – Change the number to the number of CPU's #!/usr/bin/perl print “Eating the CPUs\n”; foreach $i (1..16) { $pid = fork(); last if $pid == 0; print “Created PID $pid\n”; } while (1) { $x++; } Another form of loading all cpu’s, might not work in latest Linux bash …
Linux, Solaris, Docker, Zones – CPU hog load test Read More »
Measuring Network Latency In Solaris Use the below dtrace script (SuperPing.d) to measuring Network Latency. #!/usr/sbin/dtrace -s #pragma D option quiet #pragma D option switchrate=10hz mib:::rawipOutDatagrams /pid == $target/ { start = timestamp; } mib:::icmpInEchoReps /start/ { this->delta = (timestamp – start) / 1000; printf(“dtrace measured: %d us\n”, this->delta); @a[“\n ICMP packet delta average (us):”] …
How To Measure Network Latency With Ping And Dtrace in Solaris. Read More »
Contents 1 Solaris global and non zones performance tuning 1.1 Configuring the global zone to use FSS 1.1.1 First enable & set FSS in the global zone 1.1.2 Next configure a global zone CPU/Memory Cap 1.1.2.1 Example after the configuration change 1.2 Configure /tmp(tmpfs) system cap 1.3 reboot the system for changes to take effect …
Capture Linux network statistics Example capture output is below. collisions : 0 0 0 multicast : 2101 0 0 rx_bytes : 1079497308853 15334 12768 rx_compressed : 0 0 0 rx_crc_errors : 0 0 0 rx_dropped : 0 0 0 rx_errors : 0 0 0 rx_fifo_errors : 0 0 0 rx_frame_errors : 0 0 0 rx_length_errors …
Capturing Linux network statistics – perl script Read More »
Modify the Oracle Database Cluster process to use the FX scheduler. priocntl -s -c FX -m 60 -p 60 -i pid `pgrep -f ora_lgwr_TST` Update to include the LMS process as well for i in `pgrep -f ora_lms && pgrep -f ora_lgwr_TST`; do priocntl -s -c FX -m 60 -p 60 -i pid $i; done
Capturing Network statistics on Solaris 10 Latest nicstat utility is available here. The latest nicstat utility is compiled and might work better, below is the older perl version. The latest version is also on sourceforge Example script output Time Int rKB/s wKB/s rPk/s wPk/s rAvs wAvs %Util Sat 00:26:33 aggr1 60562.9 1521.4 41090.4 20494.1 1509.3 …
Solaris network statistics utility (only Solaris 10) Read More »
Contents 1 Resource Capping available now 1.1 CPU capping for zones 1.2 Memory capping for zones 1.3 Memory capping settings coming soon 1.4 Resource capping Helpful Tips Resource Capping available now Note: It is now much simpler to configure caps in the most recent Solaris versions click here for an example This document assumes your …
Solaris Dynamic Pool Resource Capping With Examples Read More »
This script will list the ZFS adaptive replacement cache (ARC)statistics and usages To get the script click here Another great ZFS ARC summery script is available here Output will look something like this. [root@server] /tmp # ./arc_summary.pl System Memory: Physical RAM: 64128 MB Free Memory : 58808 MB LotsFree: 999 MB ZFS Tunables (/etc/system): ARC Size: …