{"id":5339,"date":"2014-08-24T13:59:35","date_gmt":"2014-08-24T05:59:35","guid":{"rendered":"https:\/\/webhostinggeeks.com\/howto\/?p=5339"},"modified":"2023-04-28T09:48:16","modified_gmt":"2023-04-28T09:48:16","slug":"how-to-setup-collectl-linux-performance-monitoring","status":"publish","type":"post","link":"https:\/\/webhostinggeeks.com\/howto\/how-to-setup-collectl-linux-performance-monitoring\/","title":{"rendered":"How to Setup Collectl – Linux Performance Monitoring"},"content":{"rendered":"

Collectl is a linux performance monitoring tools that grabs as much detail as possible from the \/proc filesystem and it does a lot more than most other tools. Compare to sar<\/a>, collectl has some capabilities that sar does not have. Collectl can gather and post-process the performance data and also can save the performance data for later analysis. Please refer to below guide on how you can setup collectl on linux CentOS 6.5 and the sample usage of collectl :<\/p>\n

1. Install collectl. Make sure\u00a0additional repository (EPEL repository<\/a>) has been installed :
\na. For Red Hat based distro :<\/p>\n

[root@oss ~]# yum install collectl -y\n<\/pre>\n
b. For debian bas distro :<\/p>\n
[root@oss ~]# sudo apt-get install collectl -y\n<\/pre>\n
2. Display collectl command help :<\/p>\n
[root@oss ~]# collectl -h\nThis is a subset of the most common switches and even the descriptions are\nabbreviated.  To see all type 'collectl -x', to get started just type 'collectl'\n\nusage: collectl [switches]\n  -c, --count      count      collect this number of samples and exit\n  -f, --filename   file       name of directory\/file to write to\n  -i, --interval   int        collection interval in seconds [default=1]\n  -o, --options    options    misc formatting options, --showoptions for all\n                                d|D - include date in output\n                                  T - include time in output\n                                  z - turn off compression of plot files\n  -p, --playback   file       playback results from 'file' (be sure to quote\n                              if wild carded) or the shell might mess it up\n  -P, --plot                  generate output in 'plot' format\n  -s, --subsys     subsys     specify one or more subsystems [default=cdn]\n      --verbose               display output in verbose format (automatically\n                              selected when brief doesn't make sense)\n\nVarious types of help\n  -h, --help                  print this text\n  -v, --version               print version\n  -V, --showdefs              print operational defaults\n  -x, --helpextend            extended help, more details descriptions too\n  -X, --helpall               shows all help concatenated together\n\n  --showoptions               show all the options\n  --showsubsys                show all the subsystems\n  --showsubopts               show all subsystem specific options\n  --showtopopts               show --top options\n\n  --showheader                show file header that 'would be' generated\n  --showcolheaders            show column headers that 'would be' generated\n  --showslabaliases           for SLUB allocator, show non-root aliases\n  --showrootslabs             same as --showslabaliases but use 'root' names\n\nCopyright 2003-2014 Hewlett-Packard Development Company, L.P.\ncollectl may be copied only under the terms of either the Artistic License\nor the GNU General Public License, which may be found in the source kit\n<\/pre>\n
3. According to the man page, collectl identifies the following subsystems :<\/p>\n
SUMMARY SUBSYSTEMS<\/b><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Option<\/th>\nDescription<\/th>\n<\/tr>\n
b<\/i><\/td>\nBuddy information (memory fragmentation)<\/td>\n<\/tr>\n
c<\/i><\/td>\nCPU information<\/td>\n<\/tr>\n
d<\/i><\/td>\nDisk<\/td>\n<\/tr>\n
f<\/i><\/td>\nNFS information<\/td>\n<\/tr>\n
i<\/i><\/td>\ninode information<\/td>\n<\/tr>\n
j<\/i><\/td>\nInterrupts<\/td>\n<\/tr>\n
l<\/i><\/td>\nLustre<\/td>\n<\/tr>\n
m<\/i><\/td>\nMemory<\/td>\n<\/tr>\n
n<\/i><\/td>\nNetworks<\/td>\n<\/tr>\n
s<\/i><\/td>\nSockets<\/td>\n<\/tr>\n
t<\/i><\/td>\nTCP<\/td>\n<\/tr>\n
x<\/i><\/td>\nInterconnect<\/td>\n<\/tr>\n
y<\/i><\/td>\nSlabs<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
DETAIL SUBSYSTEMS<\/b><\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Option<\/th>\nDescription<\/th>\n<\/tr>\n
C<\/i><\/td>\nCPU<\/td>\n<\/tr>\n
D<\/i><\/td>\nDisk<\/td>\n<\/tr>\n
E<\/i><\/td>\nEnvironmentals via\u00a0ipmitool<\/i><\/td>\n<\/tr>\n
F<\/i><\/td>\nNFS data<\/td>\n<\/tr>\n
J<\/i><\/td>\nInterrupts<\/td>\n<\/tr>\n
M<\/i><\/td>\nMemory node data (including numa)<\/td>\n<\/tr>\n
N<\/i><\/td>\nNetworks<\/td>\n<\/tr>\n
T<\/i><\/td>\nSixty-five TCP counters (only in plot format)<\/td>\n<\/tr>\n
X<\/i><\/td>\nInterconnect<\/td>\n<\/tr>\n
Y<\/i><\/td>\nSlabs<\/td>\n<\/tr>\n
Z<\/i><\/td>\nProcesses<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
 <\/p>\n
4. Monitor cpu subsystem :<\/p>\n
[root@oss ~]# collectl -sc\nwaiting for 1 second sample...\n#< --------CPU-------->\n#cpu sys inter  ctxsw\n   0   0    34     37\n   0   0    56     40\n   0   0    38     44\n   0   0    31     35\n   0   0    36     44\n<\/pre>\n
[root@oss ~]# collectl -sC\nwaiting for 1 second sample...\n\n# SINGLE CPU STATISTICS\n#   Cpu  User Nice  Sys Wait IRQ  Soft Steal Idle\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    4    0    0     0   96\n      0     0    0    0    0    0    0     0   99\n      1     0    0    0    0    0    0     0   99\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     1    0    0    0    0    0     0   99\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n      0     0    0    0    0    0    0     0  100\n      1     0    0    0    0    0    0     0  100\n<\/pre>\n
5. Monitor memory subsystem :<\/p>\n
[root@oss ~]# collectl -sm\nwaiting for 1 second sample...\n#< -----------Memory----------->\n#Free Buff Cach Inac Slab  Map\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n   3G  19M 166M  50M  36M  34M\n<\/pre>\n
[root@oss ~]# collectl -sM\nwaiting for 1 second sample...\n\n# MEMORY STATISTICS\n# Node    Total     Used     Free     Slab   Mapped     Anon   Locked    Inact   Hit%\n     0    4095M  565208K    3543M   37112K    9408K   25492K        0   51756K 100.00\n     0    4095M  565208K    3543M   37112K    9408K   25492K        0   51756K 100.00\n     0    4095M  565208K    3543M   37108K    9408K   25492K        0   51756K 100.00\n     0    4095M  565208K    3543M   37108K    9408K   25492K        0   51756K 100.00\n     0    4095M  565208K    3543M   37108K    9408K   25492K        0   51756K 100.00\n     0    4095M  565208K    3543M   37108K    9408K   25492K        0   51760K 100.00\n     0    4095M  565208K    3543M   37036K    9408K   25492K        0   51760K 100.00\n     0    4095M  565184K    3543M   37028K    9408K   25492K        0   51760K 100.00\n     0    4095M  565184K    3543M   37028K    9408K   25492K        0   51760K 100.00\n     0    4095M  565184K    3543M   37024K    9408K   25492K        0   51760K 100.00\n     0    4095M  565184K    3543M   37024K    9408K   25492K        0   51760K 100.00\n     0    4095M  565184K    3543M   37016K    9408K   25492K        0   51760K 100.00\n     0    4095M  565168K    3543M   36972K    9408K   25492K        0   51760K 100.00\n     0    4095M  565168K    3543M   36972K    9408K   25492K        0   51760K 100.00\n     0    4095M  565168K    3543M   36972K    9408K   25492K        0   51760K 100.00\n     0    4095M  565168K    3543M   36968K    9408K   25492K        0   51760K 100.00\n     0    4095M  565160K    3543M   36932K    9408K   25492K        0   51760K 100.00\n     0    4095M  565160K    3543M   36932K    9408K   25492K        0   51760K 100.00\n     0    4095M  565160K    3543M   36932K    9408K   25492K        0   51760K 100.00\n     0    4095M  565160K    3543M   36900K    9408K   25492K        0   51760K 100.00\n     0    4095M  565160K    3543M   36900K    9408K   25492K        0   51760K 100.00\n     0    4095M  565160K    3543M   36900K    9408K   25492K        0   51760K 100.00\n<\/pre>\n
6. Monitor disk subsystem :<\/p>\n
[root@oss ~]# collectl -sd\nwaiting for 1 second sample...\n#< ----------Disks----------->\n#KBRead  Reads KBWrit Writes\n      0      0      0      0\n      0      0      0      0\n      0      0      0      0\n      0      0      0      0\n      0      0     16      3\n      0      0      0      0\n      0      0      0      0\n      0      0      0      0\n      0      0      0      0\n      0      0      0      0\n<\/pre>\n
[root@oss ~]# collectl -sD\nwaiting for 1 second sample...\n\n# DISK STATISTICS (\/sec)\n#           Pct\n#Name       KBytes Merged  IOs Size  KBytes Merged  IOs Size  RWSize  QLen  Wait SvcTim Util\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0      16      1    3    5       5     1    13     10    3\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\nsda              0      0    0    0       0      0    0    0       0     0     0      0    0\n<\/pre>\n
7. collectl like iotop :<\/p>\n
[root@oss ~]# collectl --top iokb\n<\/pre>\n
Sample output :<\/p>\n
# TOP PROCESSES sorted by iokb (counters are \/sec) 13:35:14\n# PID  User     PR  PPID THRD S   VSZ   RSS CP  SysT  UsrT Pct  AccuTime  RKB  WKB MajF MinF Command\n    1  root     20     0    0 S   18M    1M  0  0.00  0.00   0  00:01.13    0    0    0    0 \/sbin\/init\n    2  root     20     0    0 S     0     0  0  0.00  0.00   0  00:00.02    0    0    0    0 kthreadd\n    3  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.04    0    0    0    0 migration\/0\n    4  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.09    0    0    0    0 ksoftirqd\/0\n    5  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 migration\/0\n    6  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.30    0    0    0    0 watchdog\/0\n    7  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.38    0    0    0    0 migration\/1\n    8  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 migration\/1\n    9  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.22    0    0    0    0 ksoftirqd\/1\n   10  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.28    0    0    0    0 watchdog\/1\n   11  root     20     2    0 S     0     0  0  0.00  0.00   0  00:08.15    0    0    0    0 events\/0\n   12  root     20     2    0 S     0     0  1  0.00  0.00   0  01:21.61    0    0    0    0 events\/1\n   13  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 cgroup\n   14  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 khelper\n   15  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 netns\n   16  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 async\/mgr\n   17  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 pm\n   18  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.84    0    0    0    0 sync_supers\n   19  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.82    0    0    0    0 bdi-default\n   20  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 kintegrityd\/0\n   21  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 kintegrityd\/1\n   22  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.15    0    0    0    0 kblockd\/0\n<\/pre>\n
Display only top 10 processes :<\/p>\n
[root@oss ~]# collectl --top iokb,10\n<\/pre>\n
Sample output :<\/p>\n
# TOP PROCESSES sorted by iokb (counters are \/sec) 13:42:37\n# PID  User     PR  PPID THRD S   VSZ   RSS CP  SysT  UsrT Pct  AccuTime  RKB  WKB MajF MinF Command\n    1  root     20     0    0 S   18M    1M  0  0.00  0.00   0  00:01.13    0    0    0    0 \/sbin\/init\n    2  root     20     0    0 S     0     0  0  0.00  0.00   0  00:00.02    0    0    0    0 kthreadd\n    3  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.04    0    0    0    0 migration\/0\n    4  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.09    0    0    0    0 ksoftirqd\/0\n    5  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 migration\/0\n    6  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.30    0    0    0    0 watchdog\/0\n    7  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.39    0    0    0    0 migration\/1\n    8  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 migration\/1\n    9  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.22    0    0    0    0 ksoftirqd\/1\n   10  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.28    0    0    0    0 watchdog\/1\n<\/pre>\n
Learn what fields the above list can be sorted :<\/p>\n
[root@oss ~]# collectl --showtopopts\nThe following is a list of --top's sort types which apply to either\nprocess or slab data.  In some cases you may be allowed to sort\nby a field that is not part of the display if you so desire\n\nTOP PROCESS SORT FIELDS\n\nMemory\n  vsz    virtual memory\n  rss    resident (physical) memory\n\nTime\n  syst   system time\n  usrt   user time\n  time   total time\n  accum  accumulated time\n\nI\/O\n  rkb    KB read\n  wkb    KB written\n  iokb   total I\/O KB\n\n  rkbc   KB read from pagecache\n  wkbc   KB written to pagecache\n  iokbc  total pagecacge I\/O\n  ioall  total I\/O KB (iokb+iokbc)\n\n  rsys   read system calls\n  wsys   write system calls\n  iosys  total system calls\n\n  iocncl Cancelled write bytes\n\nPage Faults\n  majf   major page faults\n  minf   minor page faults\n  flt    total page faults\n\nContext Switches\n  vctx   volunary context switches\n  nctx   non-voluntary context switches\n\nMiscellaneous (best when used with --procfilt)\n  cpu    cpu number\n  pid    process pid\n  thread total process threads (not counting main)\n\nTOP SLAB SORT FIELDS\n\n  numobj    total number of slab objects\n  actobj    active slab objects\n  objsize   sizes of slab objects\n  numslab   number of slabs\n  objslab   number of objects in a slab\n  totsize   total memory sizes taken by slabs\n  totchg    change in memory sizes\n  totpct    percent change in memory sizes\n  name      slab names\n<\/pre>\n
8. collectl like top :<\/p>\n
[root@oss ~]# collectl --top\n<\/pre>\n
Sample output :<\/p>\n
# TOP PROCESSES sorted by time (counters are \/sec) 13:45:00\n# PID  User     PR  PPID THRD S   VSZ   RSS CP  SysT  UsrT Pct  AccuTime  RKB  WKB MajF MinF Command\n 3266  root     20  2488    0 R  160M   21M  1  0.00  0.05   5  00:00.70    0    0    0   83 \/usr\/bin\/perl\n    1  root     20     0    0 S   18M    1M  0  0.00  0.00   0  00:01.13    0    0    0    0 \/sbin\/init\n    2  root     20     0    0 S     0     0  0  0.00  0.00   0  00:00.02    0    0    0    0 kthreadd\n    3  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.04    0    0    0    0 migration\/0\n    4  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.09    0    0    0    0 ksoftirqd\/0\n    5  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 migration\/0\n    6  root     RT     2    0 S     0     0  0  0.00  0.00   0  00:00.31    0    0    0    0 watchdog\/0\n    7  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.39    0    0    0    0 migration\/1\n    8  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 migration\/1\n    9  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.22    0    0    0    0 ksoftirqd\/1\n   10  root     RT     2    0 S     0     0  1  0.00  0.00   0  00:00.28    0    0    0    0 watchdog\/1\n   11  root     20     2    0 S     0     0  0  0.00  0.00   0  00:08.18    0    0    0    0 events\/0\n   12  root     20     2    0 S     0     0  1  0.00  0.00   0  01:21.97    0    0    0    0 events\/1\n   13  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 cgroup\n   14  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 khelper\n   15  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 netns\n   16  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 async\/mgr\n   17  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 pm\n   18  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.84    0    0    0    0 sync_supers\n   19  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.83    0    0    0    0 bdi-default\n   20  root     20     2    0 S     0     0  0  0.00  0.00   0  00:00.00    0    0    0    0 kintegrityd\/0\n   21  root     20     2    0 S     0     0  1  0.00  0.00   0  00:00.00    0    0    0    0 kintegrityd\/1\n<\/pre>\n
More advance options can be found in the official collectl documentation<\/a> :<\/p>\n","protected":false},"excerpt":{"rendered":"
Collectl is a linux performance monitoring tools that grabs as much detail as possible from the \/proc filesystem and it does a lot more than most other tools. Compare to…<\/p>\n","protected":false},"author":6,"featured_media":6280,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"wds_primary_category":0,"footnotes":""},"categories":[2055],"tags":[1256,1286,2085],"class_list":["post-5339","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-centos","tag-centos-6-5","tag-collectl","tag-performance"],"_links":{"self":[{"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/posts\/5339","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/comments?post=5339"}],"version-history":[{"count":0,"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/posts\/5339\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/media\/6280"}],"wp:attachment":[{"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/media?parent=5339"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/categories?post=5339"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/webhostinggeeks.com\/howto\/wp-json\/wp\/v2\/tags?post=5339"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}