collect man page on OSF1

Man page or keyword search:  
man Server   12896 pages
apropos Keyword Search (all sections)
Output format
OSF1 logo
[printable version]

collect(8)							    collect(8)

NAME
       collect - Collects data that describes the current system status

SYNOPSIS
       /usr/sbin/collect - [-a | -c | -F | -h | -i I: [PI] | -l | -nNum | -S |
       -t | -T | -v| -V]

       /usr/sbin/collect [-C start_time,end_time]

       /usr/sbin/collect [-D device1, [device2, ... deviceN]]

       /usr/sbin/collect [-e | -s [pmdtlncfqyha]]

       /usr/sbin/collect [-f file [options]]

       /usr/sbin/collect [-H h d w m time[,how_long]]

       /usr/sbin/collect [-L group]

       /usr/sbin/collect [-M suspend_value,resume_value]

       /usr/sbin/collect [-o [tmfnzlq]]

       /usr/sbin/collect [-pfile1] [-p file2] [-p fileN]

       /usr/sbin/collect [-p collect_datafile [-f output_datafile]]

       /usr/sbin/collect [-P pid1, ... pidN | [P pid1, ...  pidN]  |  [C  com‐
       mand1, ... commandN] | [U user/UID1, ... user/UIDN]]

       /usr/sbin/collect [-R NumberUnit...]

       /usr/sbin/collect [-W NumberUnit ...]

OPTIONS
       Directs the output from the collect utility to stdout, which is usually
       the screen or window from which the collect command was entered.	  This
       is  the	default behavior if no data collection file is specified using
       the -f option.  Simultaneously displays	collect	 data  on  the	screen
       (stdout)	 and  records  the data in a file when the -f option is speci‐
       fied.  Specifies which AdvFS objects are included for data  collection.
       The  AdvFS  object  can be a domain name or the domain#fileset combina‐
       tion. The domain name tells the collect utility to collect  volume  I/O
       queue data for the domain. The domain#fileset combination instructs the
       collect utility to collect

       fileset vnode operation data for the fileset specified.	The  following
       is      an      example	    command	line:	  #	collect	    -A
       usr_domain,root_domain#root,usr_domain#usr,var_domain

	      This command line instructs collect to gather volume  I/O	 queue
	      data for the usr_domain and var_domain domains, and gather file‐
	      set  vnode  operation  data   for	  the	root_domain#root   and
	      usr_domain#usr  filesets.	  In  a TruCluster Server environment,
	      directs collect to gather local and remote  I/O  access  statis‐
	      tics for cluster disk and tape storage devices.

	      During  collection,  the	-c option, if specified, is ignored if
	      collect is run on a system that is not a cluster member.	During
	      playback,	 the  -c  option  must be specified to display cluster
	      storage device I/O information  from an input datafile that con‐
	      tains cluster data.

	      See  also	 the option descriptions for the disk and tape subsys‐
	      tems.  Extracts a series of samples from a file according to the
	      specified start time and end time for the series.	 The format of
	      the time string is:

	      [+]Year:Month:Day:Hour:Minute:Second

	      For example: +2010:11:18:23:30:00.

	      Every time string field is optional, except the last (the Second
	      field).  If  any	of  the optional fields are not specified, the
	      corresponding values from the Start or End time are used.

	      The optional + (plus) sign at the beginning of the  time	string
	      indicates	 that  time  is	 relative to the beginning of the data
	      collection period. If + is not specified, the  -C	 option	 indi‐
	      cates absolute time.

	      If  the start-time argument is omitted, the start of the collec‐
	      tion period is used. If the end-time argument  is	 omitted,  the
	      end of the collection period is used.

	      Use the -p option to specify the data file from which to extract
	      the samples. For example:

	      # collect -C 14:37:34,14:38:40 -p	 input_file.cgz	 Prints	 debug
	      information  to  stdout.	Specifies which disks are included for
	      data collection, using the device special filename of the disks,
	      such  as	dsk3  for SCSI disk number 3. Running collect with the
	      -sd option will produce a list of all the disks  known  to  col‐
	      lect.  You  can also obtain a complete list of disk devices from
	      the  device  directories	under  /dev,  such  as	/dev/disk   or
	      /dev/tape,  but  collect	may not be aware of all these devices.
	      Use the hwmgr command to identify devices and obtain device name
	      data.  See  the  hwmgr(8)	 reference page for information on the
	      command options.

	      You can also use regular	expressions  to	 specify  a  group  of
	      disks.  For  example  "dsk.*"   selects  all  disks (enclose the
	      expression in either single or double quotes).  For  information
	      on  regular  expressions, refer to the grep(1) reference page or
	      the Programming Support Tools guide.

	      When the -c option  is  specified	 and  the  disk	 subsystem  is
	      included	in data collection, the cluster disk device I/O infor‐
	      mation follows the "DISK Statistics" section in  the  output  of
	      the  collect  command.  See d -disk for sample output.  Excludes
	      the specified subsystems from the data collection and  playback.
	      Do  not  enter  a space between letters when specifying options.
	      For example, the following command specifies that only  the  CPU
	      and file system data be excluded: # collect -e cf

	      The  option  letters  map to the following subsystems: Specifies
	      that process data (as shown in the Process Statistics (RSS & VSZ
	      in  MBytes) section of the output report) are excluded from data
	      collection. When included, this data appears similar to the fol‐
	      lowing in the output from the collect command:
	       PID User %CPU RSS VSZ   UsrTim  SysTim IBk OBk Maj Min  Command
		0  root	  2.1  12M  342M    0.00    0.01  0    0   0   0  kern
	      idle .  .	 .  Specifies that memory data (as shown in the MEMORY
	      STATISTICS  section of the output report) are excluded from data
	      collection. When included, this data appears similar to the fol‐
	      lowing  in the output from the collect command: # MEMORY STATIS‐
	      TICS  (<--------	 MegaBytes   ------->	<---------   Pages/sec
	      ------->)
	       Free  Swap  Act	InAc Wire UBC  PI PO Zer Re COW SW HIT PP ALL
		135	 1   44	   22	40  36	 0  0	7  0   0  0   0	 0   0
	      Specifies that the disk data (as shown in	 the  DISK  Statistics
	      section of the output report) are excluded from data collection.
	      When included, this data appears similar to the following in the
	      output from the collect command:
	       DISK Statistics
	       DSK NAME B/T/L R/S RKB/S W/S  WKB/S  AVS	 AVW ACTQ WTQ %BSY
		 0 dsk0 0/0/0  53   431	  0	 5 8.49 0.00 0.46 0.00 43.09

	      When  the	 -c  option  is	 specified  and	 the disk subsystem is
	      included in data collection, the cluster disk device I/O	infor‐
	      mation  follows  the  "DISK Statistics" section in the output of
	      the collect command and appears  similar	to  the	 following:  #
	      CDISK  CLUSTER DISK I/O Statistics #Cnt MembID TargetDsk DrdDev‐
	      Type  R/S RKB/S  W/S WKB/S
		 0	1      dsk0	    ss	  0	0    0	   0
		 1	1      dsk1	    sd	  0	0    3	  35
		 2	1      dsk2	    sd	  0	0    0	   0
		.
		.
		.
		18	2      dsk0	    co	  0	0    0	   0
		19	2      dsk1	    co	  0	0    0	   0
		20	2      dsk2	    sd	  0	0    0	   0
		.
		.
		.  Specifies that tape device data is excluded from data  col‐
	      lection.

	      When  the	 -c  option  is	 specified  and	 the tape subsystem is
	      included in data collection, the cluster tape device I/O	infor‐
	      mation  follows  the  "TAPE Statistics" section in the output of
	      the collect command and appears  similar	to  the	 following:  #
	      CTAPE CLUSTER TAPE I/O Statistics #Cnt MembID TargetTape DrdDev‐
	      Type  R/S RKB/S  W/S WKB/S
	       0      1	     tape0	   ss  496   248    0	  0
	       1      1	     tape1	   ss	 0     0    0	  0
	       2      2	     tape0	   cs	 0     0    0	  0
	       3      2	     tape1	   cs	 0     0    0	  0  Specifies
	      that  LSM	 volume	 data  is  excluded from data collection. When
	      included, this data appears similar to the following in the out‐
	      put  from the collect command, one volume at a time: #LSM Volume
	      Statistics #VOL  NAME   R/S  RKB/S  RAVS	W/S  WKB/S   WAVS
		1 rootvol    0	   0   0.00    0    12	 45.62 Specifies  that
	      the  network data (as shown in the Network Statistics section of
	      the output report)  are  excluded	 from  data  collection.  When
	      included,	 this  data  appears as follows in the output from the
	      collect command: #  Network Statistics  #Cnt  Name  Inpck	 InErr
	      Outpck OutErr Coll IKB OKB %BW
		 2   tu0     89	    0	   2	  0    0  10   0   0 Specifies
	      that the CPU data (as shown in the CPU SUMMARY and  CPU  STATIS‐
	      TICS  sections of the output report) are excluded from data col‐
	      lection. When included, this data appears as follows in the out‐
	      put  from	 the  collect  command: CPU SUMMARY USER SYS IDLE WAIT
	      INTR SYSC	 CS RUNQ AVG5 AVG30 AVG60 FORK VFORK
		13  16	 71    0  149  492 725	   0  0.13   0.05   0.01  0.30
	      0.00 SINGLE CPU STATISTICS CPU USER SYS IDLE WAIT
		0   13	16   71	   0 Specifies that file system data (as shown
	      in the FileSystem Statistics section of the output  report)  are
	      excluded	from data collection. When included, this data appears
	      as follows in the output from the collect command: #  FileSystem
	      Statistics # FS	    Filesystem		Capacity   Free
		 0	 root_domain#root	  128	  30
		 1		    /proc	    0	   0
		 2	 usr_domain#usr		  700	 147
		 3	  usr_domain#var	    700	    147 Specifies that
	      terminal data (as shown in the TTY  Statistics  section  of  the
	      output report) are excluded from data collection. When included,
	      this data appears as follows in the output from the collect com‐
	      mand: #  TTY Statistics #	 In    Out    Can    Raw
		  3   1489	0      3 Specifies that the record headers are
	      excluded from data collection. When included, this data  appears
	      as   follows   in	  the	output	 from	the  collect  command:
	      ################################################################
	      # OSF1 pauli.zso.cpqcorp.net V5.1 732 DEC4100 2/531MHz/512MB   #
	      # HOST..pauli.zso.cpqcorp.net Started..Tue Dec 5 11:48:11 2000 #
	      # Seconds........976045691				     #
	      #								     #
	      # CPU FAMILY....21164 (EV5 core)	 CPU ID....EV5.6 (21164A)    #
	      # CPU EXTENSIONS..BYTE					     #
	      # PLATFORM NAME...DEC4100 CPU	     SPEED......531 MHz	     #
	      # SWAP SIZE.......1005 MB		     Physical Mem...512 MB   #
	      # NUM CPUS........2 NUM DISKS......4			     #
	      # NUM LANS........4 NUM		     FSYS.......4	     #
	      # MAX MQUEUES.....64		     NUM TAPES......0	     #
	      # INTERVAL........1.00		     PROC_INTERVAL..1.00     #
	      # UBCMAXPERCENT...100		     UBCMINPERCENT..10	     #
	      # MAXUSERS........512		     MAXUPRC........64	     #
	      # Delay_WBuffers..0		     LSM Volumes....0	     #
	      ################################################################

	      #### RECORD    1 (976045693:45) (Tue Dec	5 11:48:13 2000)  ####

	      When the -c option is specified and record headers are included,
	      the  header  data displayed by collect includes the cluster name
	      and number of configured	members	 for  datafiles	 that  contain
	      cluster  data.   Specifies  that	AdvFS data  are excluded  from
	      data collection. The AdvFS data shows the Volume I/O queue  sta‐
	      tistics and Fset vnode operations for AdvFS. When included, this
	      data appears as follows in the output  from the collect command:
	      # Volume I/O queue statistics for AdvFS

	      #	 domain.vol	 rd  wr rg arg wg awg blk flsh wlz sms rlz con
	      dev root_dmn.001	0  0  0	  0  0	 0   0	  0  17	   0	0    0
	      0 usr_dmn.001   0	 0  0	0  0   0   0	0   1	1   0	0   0

	      # Fset vnode operations for AdvFS

	      #	 fileset      lkup crt geta read writ fsnc dsnc rm mv rdir mkd
	      rmd lnk root_dmn#root 11	 0   20	   7	0    0	  0   0	 0   0
	      0	   0	0  usr_dmn#us	 0   0	  0    0    0	 0    0	  0  0
	      0	  0   0	  0 usr_dmn#va	  0   0	   0	0    1	  0	0    0
	      0	  0   0	  0   0

	      When  the	 exclusion  option  is	used,  only  the RECORD N line
	      appears.

	      See also the -s option, which is used to specify subsystems that
	      must be included in data collection.  Records data in the speci‐
	      fied file. The argument is  a  path  name	 to  a	file  such  as
	      /usr/users/collectdata/nov13.  By	 default,  the collect command
	      creates a compressed file and appends a extension	 to  the  file
	      name that you specify. For example, the file nov13 is created as
	      nov13.cgz. (See the -o option if you want to  create  an	uncom‐
	      pressed file.)

	      The  collect -f command option creates a binary format file. Use
	      the -p option if you want to replay the contents of the file.

	      See also the -a option,  which  enables  you  to	simultaneously
	      direct  the  output  from the collect utility to stdout (usually
	      the terminal from which the collect utility is invoked). You can
	      also  specify  other data collection options with the -f option,
	      such as -s or -n, to control what information is recorded in the
	      file.

	      See the section on Historical Mode Usage in the Description sec‐
	      tion for information about restrictions and file name  construc‐
	      tion when using collect in historical mode.  Displays or records
	      full process information lines, including those longer  than  80
	      columns.	The process priorities are shown, with the RSS and VSZ
	      values represented in kilobytes. The following is	 example  out‐
	      put,  except  that  here	the  column  widths have been manually
	      adjusted	to  show   the	 example   output:   ####   RECORD   1
	      (976045693:45) (Tue Dec 5 11:48:13 2000) ####

	      Process  Statistics  (RSS & VSZ in KBytes) PID PPID Usr %CPU RSS
	      VSZ UsrTim SysTim Pri IBk OBk Maj Min Command
	       0  0  root  2.4 1984 3744 0.00	0.02	0  0   0   0   0  kern
	      idl
	       1   0   root   0.0    96	  480  0.00   0.00   44	 0   0	 0   0
	      init .  .	 .

	      Compare the preceding output  to	the  example  output  for  the
	      Process  Statistics  report  section, shown in the entry for the
	      -ep option.  Display a usage summary (help) for the collect com‐
	      mand line options.  Runs the collect utility in historical mode.
	      The how_long argument defines the length of time that  the  logs
	      are  preserved.  The how_long argument is optional and if you do
	      not specify it the log preservation default is one week.

	      Time variables are indicated as follows: MM  -  Minute,  in  the
	      range 0-59.  HH - Hour, in the range 0-24.  WD - Weekday, in the
	      range 0-6, with 0 representing Sunday.  MD - Day of  the	month,
	      in  the range 1-31.  The following values for time can be speci‐
	      fied for each argument: An  hourly  rollover  at	the  specified
	      minute.  A  value	 of  -Hh3 will roll over the collect log every
	      hour at three minutes past the hour. For	example:  0:03,	 1:03,
	      2:03,  and  3:03.	  A  daily  rollover at the specified hour and
	      minute in 24-hour time format. For example, a value  of  -Hd14:2
	      will  roll  over	the collect log every day at system time 14:02
	      (2:02 PM).  A weekly rollover at the  specified  day,  hour  and
	      minute in seven-day, 24-hour time format. A value of zero (0) in
	      the day  field  represents  Sunday.  For	example,  a  value  of
	      -Hw1@10:25  will roll over the collect log every Monday at 10:25
	      (10:25 AM).  A monthly rollover at the specified date, hour, and
	      minute  in  31-day  24-hour time format. For example, a value of
	      -Hm3@21:15 will roll over the collect log every third day of the
	      month at 21:15 (9:15 PM).

	      Declare  the value of time by specifying day and week values for
	      how_long. For example -Hd14:12,2d5w will roll over the log every
	      day at 14:12 (2:12 PM) and keep the log for 2 days and 5 weeks.

	      See the section on Historical Mode Usage in the Description sec‐
	      tion for information about restrictions and file name  construc‐
	      tion  when  using	 collect in historical mode.  Specifies a time
	      value in seconds for the interval (I) and, optionally, the  time
	      value  for    the	 process  subsystem data collection interval (
	      PI). This enables you to control the rate at which data is  col‐
	      lected from subsystems.  Floating-point values are permitted.

	      When  you	 use this option, the initialization message echoed by
	      the collect utility is updated to confirm the  value  of	I,  as
	      follows: # collect -i 2:8 Initializing (2.0 seconds)(float OK) #
	      collect -i 5:12 PROC_INTERVAL must be evenly divisible by INTER‐
	      VAL

	      Note  that  in the second command, an error message is displayed
	      because the value of PI must always be evenly divisible  by  the
	      value of I.

	      The capabilities of the machine and the number of subsystems for
	      which records are requested affect the collect command's ability
	      to  return  data.	 Sub-second  sampling  intervals  and complete
	      record requests will impose limits on data collection  for  many
	      machines.	  Looks	 for the last valid record and prints it. This
	      is primarily used by the graphical interface to get  the	ending
	      time of the collection period.  Collects data only for the spec‐
	      ified LSM disk group, listed  in	/dev/vol.  To  make  the  name
	      unique,  the  format is the disk group name, for example: my_dg.
	      Regular expressions can be specified  to	select	the  LSM  disk
	      group.   Monitors	 free disk space.  collect suspends writing to
	      disk when free disk space falls below a declared	threshold  and
	      resumes when free space rises above the threshold.

	      In the following example, collect suspends disk writes when free
	      disk space falls below 250 megabytes, and resumes	 writing  when
	      free  disk space rises above 300 megabytes: # collect -M 250,300
	      Selects only top Num processes, where Num is  an	integer.  This
	      option  is  useful  with	the  -S	 sorting option.  Options that
	      enable you to control the data collection procedure: Show	 abso‐
	      lute  system  and	 user  time  (T	 in  data recorded for the the
	      process subsystem), the way the ps command does. The default  is
	      to  show	a  one	second normalized delta since the last sample,
	      thus making graphs of these time values more useful.  Show  8192
	      byte  pages instead of megabytes for absolute memory values.  Do
	      not prompt before overwriting an existing output file.   Do  not
	      allow  the  collect  utility to set high scheduling priority for
	      itself using the	nice  command.	 Do  not  write	 a  compressed
	      (zipped) output file.  Prevents the collect utility from locking
	      its pages into  memory.	Causes	the  collect  utility  to  use
	      instantaneously  measured	 queue	lengths, instead of calculated
	      averages.	 When an existing collect_datafile is specified alone,
	      the  collect utility plays back the contents of the file to std‐
	      out (usually the terminal window from which the collect  command
	      was entered). You use options such as -e to filter the data read
	      from the collect_datafile. As the file contents will  be	large,
	      you  may	want to pipe the output to the more command or use the
	      grep command to search for specific data items.

	      To convert data files created using previous  versions  of  col‐
	      lect,  use  the -f option to specify an output_datafile.	Speci‐
	      fies process identifiers for which data should be collected. The
	      following	 process  identifiers  can  be specified: Collect data
	      only for processes in list. Specify  the	percent	 sign  (%)  to
	      include  the process for the collect command.  Collect data only
	      for processes whose parent PID (PPID) is specified, or that  are
	      members  of  a  process  group (PGID) with the same ID.  Collect
	      data only for processes whose process names contain  the	speci‐
	      fied  string.  This  can	be  a  partial	string, but must match
	      exactly. Regular expressions are not allowed.  Collect data only
	      for  processes  owned  by	 the specified users. User identifiers
	      (UIDs)  can  be  used  in	 place	of  the	 user  name.  See  the
	      /etc/passwd file for a list of user account names and associated
	      UIDs.  Specify the duration of data collection.  Either  of  the
	      following	 formats  can  be specified: The value of Number is an
	      integer. The value of Unit is one of the following: w  -	weeks,
	      such  as	4w for four weeks.  d - days, such as 2d for two days.
	      h - hours, such as 12h for twelve hours.	m - minutes,  such  as
	      30m  for	thirty	minutes.  s - seconds, such as 45s for 45 sec‐
	      onds.  Any valid combination of times can be  entered,  such  as
	      4w2d6h45m20s.  The same time format described for the -C option,
	      except that a plus sign (+) indicates the value is  relative  to
	      the  current time. Without a plus sign, the value is an absolute
	      time at which the data collection period	should	end.   Include
	      the  specified subsystems in data collection and playback, which
	      can be: Proc, Mem, Disk, Tape, Lsm, Net, Cpu,  Filesys,  mQueue,
	      ttY,  Header,  and  AdvFS). The option letters (p m d...) map to
	      these subsystems and are described under the entry  for  the  -e
	      option.

	      Do  not  enter  a space between letters when specifying options.
	      For example, the following command specifies that only  the  CPU
	      and file system data are included: # collect -s cf

	      If all specified subsystems are unavailable on the local system,
	      only a RECORD N header will be displayed. The following  example
	      shows  what  happens  when  t  (tape)  is specified, but no tape
	      device exists on the system: # collect -s t .  .	#### RECORD  4
	      (943046239:0)  (Fri  Feb	16  16:17:19 2001) #### #### RECORD  5
	      (943046249:0) (Fri Feb 16 16:17:29 2001) #### .  .   Sorts  pro‐
	      cesses  according	 to their %CPU usage (percentage of processing
	      time used).  Prefixes a tag (or marker) to  all  data  lines  to
	      facilitate manipulation of data by scripts.  Specifies only that
	      total disk and tape throughput be recorded or displayed  as  the
	      Sum MB/sec (megabytes per second). All other subsystems are des‐
	      elected.	Enables verbose mode, listing the devices attached  to
	      the system as shown in the following sample output: % collect -v
	      No objects found of type hardware/tape

		      found 4 Disks, 0 Tapes
		      found CPU 0 at slot [0]
		      found CPU 1 at slot [1]
		      max_procs = 16384
		      SAMPLE: 0
		      Initializing (10.0 seconds) ...	Displays  the  Collect
	      executable  version,  and, if used with the -p option, also dis‐
	      plays the version of the data file.   Specifies  that  data  are
	      written  to  the output file at least once per period stipulated
	      in the NumberUnit argument.

	      The argument is a compound of Number,  an	 integer  representing
	      the  amount  of  the  given time unit, and Unit, which is one or
	      more of the time options (w, d, h, and m), such as in  the  fol‐
	      lowing examples:
	       # collect -H -W 1h -f filename
	       # collect -H -W 1h30m -f filename

	      The  former  command writes data to disk once per hour while the
	      latter writes data to disk every 90  minutes.  Writing  to  disk
	      requires	using the file option (f) specifying the file in which
	      to record the data.

DESCRIPTION
       The collect utility is a system monitoring tool that  records  or  dis‐
       plays  specific	operating system data. Any set of the subsystems, such
       as file systems, message queue, tty, or header can be  included	in  or
       excluded from data collection. You can display data at the terminal, or
       store it in either a compressed or uncompressed data file.  Data	 files
       can  be	read  and manipulated from the command line, or through use of
       command scripts.

       To ensure that collect delivers reliable	 statistics  it	 locks	itself
       into  memory  using  the	 page locking function plock(), and by default
       cannot be swapped out by the system. It also raises its priority	 using
       the  priority  function nice(). However, these measures should not have
       any impact on a system under normal load, and they should have  only  a
       minimal impact on a system under extremely high load.  If required, you
       can disable page locking using the -ol command option and disable  col‐
       lect command's priority setting using the -on command option.

       Some  collect  operations  use  kernel  data that is only accessible to
       root. System administration practice should not involve lengthy	opera‐
       tions  as  root, therefore collect is installed with permissions set as
       04750. This setting allows group (typically system) members to run col‐
       lect  with  owner  setuid permissions. If this is inappropriate in your
       environment, you can reset permissions to fit your needs.

   Automatic Starting on a Reboot
       You can configure  collect  to  start  automatically  when  the	system
       reboots.	 This  is  particularly useful for continuous monitoring.To do
       this, use the rcmgr command with the set	 operation  to	configure  the
       following   values   in	the  /etc/rc.config*  file:  %rcmgr  set  COL‐
       LECT_AUTORUN 1

       A value of 1 sets collect to automatically start on reboot. A value  of
       0 (the default) causes

       collect to not start on reboot.

       % rcmgr set COLLECT_ARGS ""

       A  null	value causes collect to start with the default values (command
       options) of:
	-i60,120 -f /var/adm/collect.dated/collect -W 1h -M 10,15

       You can select other values.  %rcmgr set COLLECT_COMPRESSION 1

       A value of 1 sets compression on. A value of 0 sets compression off.

       See the rcmgr(8) reference page for more information.

   Playing Back Multiple Data Files
       Use the collect utility with the -p option to read multiple binary data
       files  and  play them back as one stream, with monotonically increasing
       sample numbers. You can also combine multiple binary input  files  into
       one binary output file, by using the -p option with the input files and
       the -f option with the output file.

       The collect utility will combine input  files  in  whatever  order  you
       specify on the command line. This means that the input files must be in
       strict chronological order if you want to do further processing of  the
       combined output file. You can also combine binary input files from dif‐
       ferent systems, made at different times, with differing subsets of sub‐
       systems	for  which  data has been collected. Filtering options such as
       -e, -s, -P, and -D can be used with this function.

   Normalization of Data
       Where appropriate, data is presented in units per second. For  example,
       disk data such as kilobytes transferred, or the number of transfers, is
       always normalized for 1 second. This happens no matter what time inter‐
       val  is	chosen.	  The  same  is true for the following data items: CPU
       interrupts, system calls, and  context  switches.   Memory  pages  out,
       pages  in,  pages zeroed, pages reactivated, and pages copied on write.
       Network packets in, packets out, and collisions.	 Process user and sys‐
       tem time consumed.

       Other  data is recorded as a snapshot value. Examples of this are: free
       memory pages, CPU states, disk queue lengths, and process memory.

   The Data Collection Interval
       A collection interval can be specified using the -i option followed  by
       an  integer,  optionally	 followed (without spaces) by a comma or colon
       and another integer. If the optional second integer is given, this is a
       separate time interval which applies only to the process subsystem. The
       process interval must be a multiple of the regular interval. Collecting
       process	information  is	 more  taxing on system resources than are the
       other subsystems and is not generally needed  at	 the  same  frequency.
       Process	data  also takes up most space in the binary data file. Gener‐
       ally, specifying	 a  process  interval  greater	than  1	 significantly
       decreases the load on the system being monitored.

   Specifying What Data to Collect
       Use  the	 -s  (select) option to select subsystems for inclusion in the
       data collection, or use the -e (exclude) option to  exclude  subsystems
       from the data collection.

       When you are collecting process data, use the -S (sort) and - n X (num‐
       ber) options to sort data by percentage of CPU usage and to save only X
       processes.  Target  specific  processes	using the -Plist option, where
       list is a list of process identifiers, comma-separated without blanks.

       If there are many (greater than 100)  disks  connected  to  the	system
       being  monitored,  use  the  -D	option	to monitor a particular set of
       disks.

   Data Compression
       collect reads and writes compressed datafiles in	 gnuzip	 format.  Com‐
       pressed	output is enabled by default but can be disabled using the -oz
       option. The extension is appended to the output	filename,  unless  you
       specify	the  -oz  command option. You can compress older, uncompressed
       datafiles using the gzip command and the resulting files can be read by
       collect in their compressed form.

       Compression  during  collection	should not generate any additional CPU
       load. Because compression uses buffers and therefore does not write  to
       disk  after  every  sample, it makes fewer system calls and its overall
       impact is negligible. However, because the output is buffered there  is
       one possible drawback. If collect terminates abnormally (perhaps due to
       a system crash) more data samples will be lost than if  compression  is
       not used. This should not be an important consideration for most users,
       as you can specify how often data is written to the disk.

   Specifying a Time Range from a Playback File
       You can select samples from the total period of the time that data col‐
       lection ran. Use the -C option to specify a start time and, optionally,
       an end time. The format is as follows:

       [+]Year:Month:Day:Hour:Minute:Second.

       The plus sign (+) indicates that the time should be interpreted as rel‐
       ative  to  the beginning of the collection period. If any of the fields
       are excluded from the string, the corresponding values from  the	 start
       time  are used in their place as the time value is parsed from right to
       left. Thus, field one is interpreted as Second, field two (if there  is
       one),  as  Minute,  and so on. For example, if the collection period is
       from February 16, 2001, 16:44:03 to February 16,	 2001,	16:54:55,  and
       you  wish  to  extract  one  minute, all but minutes and seconds can be
       omitted from  the  command  option:  -C46:00,47:00  (from  16:46:00  to
       16:47:00). However, if the collection ran overnight, it is necessary to
       specify the day as well. For example, when the period is	 February  16,
       16:44  to  February  17, 9:30, enter the following command to specify a
       time range from 23:00 to 1:00: # -C16:23:00:00,17:1:00:00

   Historical Mode Usage
       When collect is run in historical mode (the -H option) it constructs  a
       more  complex  file name based on the parameter you specify with the -f
       option. In addition to adding the .cgz extension, collect adds user and
       date information.

       There  are two modes for file name construction, user's mode and script
       mode. If you run collect directly, collect will expand the file name to
       this format: filename_user@date.cgz

       For  example,  if  you  specify a file name of collect.dat at midday on
       June  24th,  collect  will  construct  this  full   file	  name:	  col‐
       lect.dat_user@24-Jun-12:26:54.cgz

       If instead you run collect from a script, the name construction will be
       of this form: collect.dat_init@24-Jun-12:26:54.cgz

       If you are running more than one instance of collect, there is a possi‐
       bility  of  creating more than one file simultaneously. If this occurs,
       collect manages the potential name collsion  by	appending  incremental
       numbers	to  the	 files.	 For instance: data_user@24-Jun-12:36:01-1.cgz
       data_user@24-Jun-12:36:01-2.cgz data_user@24-Jun-12:36:01-3.cgz

       When using the -f and -H options together on either a  clustered	 or  a
       non-clustered  system,  the  directory  var/adm/collect.dated  must  be
       present as a symbolic link.

   General Command Options
       The following command options are useful: Use the -a option to  display
       simultaneous  text  (ASCII)  output to the screen while collecting to a
       file.  Use the -t option to prefix each data line with  a  unique  tag.
       This makes it easier for your scripts to find and to extract data. Tags
       are superfluous if you use the perl script cfilt.  Use the -T option to
       shut  off collection for all subsystems except disk, and only display a
       total megabytes per second (MB/sec) across all disks in the system. Use
       the  -s option with the -T option to override this behavior and collect
       data for other subsystems.  Use the -R  to  terminate  data  collection
       after a specified amount of time.

       All  flags  that can reasonably be applied to both collection and play‐
       back will work. The -Plist filter option used  during  collection  col‐
       lects  data only for the processes you specify. During playback it dis‐
       plays only data for the corresponding processes. To save space  in  the
       binary  data file, you can limit your collection to specific processes,
       specific disks, or specific subsystems. However, if you want to look at
       volumes	of data and select different chunks at a time, you should col‐
       lect everything and later use the filter options to select  data	 items
       during playback.

   Disk Statistics
       Note  that under certain circumstances the data provided under the Disk
       Statistics section of the output report might be only approximate.  For
       older  releases of collect, some data fields were zero and data in some
       fields could be inaccurate under certain circumstances.

   Data Conversion and Filtering
       collect automatically reads older datafile versions when	 playing  back
       files.

       You  can	 convert an older collect version datafile to the current ver‐
       sion using the -p collect_datafile option with the -f fileoption.  Dur‐
       ing  conversion	you  can  use most command options to extract specific
       data from the input collect_datafile. For example: Use the  -s  and  -e
       options	to  select  data only from particular subsystems.  Use the -nX
       and -S options to take only X processes and sort	 them  by  CPU	usage.
       Use  the -D option to select disks and the -L option to select LSM vol‐
       umes.  Use the -P, -PC, -PU, and -PP options to select processes	 based
       on  their  identifiers.	Use the -C option to extract data according to
       specified start and stop times.

   Cluster Storage Device I/O Statistics
       When the -c option is specified in  a  TruCluster  Server  environment,
       collect	gathers	 local	and  remote I/O access statistics for disk and
       tape devices, as seen by the DRD (Device	 Request  Dispatcher)  cluster
       subsystem.

       Each  line  in  the  command's  cluster storage device I/O report shows
       information for I/O dispatched by the DRD on the specified cluster mem‐
       ber (whose member ID is shown) to the specified device. Either the disk
       or tape subsystems (or both) must have been included  in	 data  collec‐
       tion.

       Changes	in  cluster  membership	 state are detected by collect and its
       output is adjusted accordingly.

       The -D option can also be used to monitor  a  subset  of	 cluster  disk
       devices.

       See  also  the drdmgr(8) reference page and the Cluster Technical Over‐
       view manual  for more information on cluster storage  devices  and  how
       they are served in a cluster.

   DataFields
       The  following  table provides definitions for the data fields that you
       might see in any output from collect.

       ────────────────────────────────────────────────────────────────
       Data Field   Description
       ────────────────────────────────────────────────────────────────
       Process Section
       PID	    The process ID.
       User	    The user name.
       %CPU	    The percent of the CPU(s) the process is currently
		    (more or less) using.
       RSS	    Resident   Set   Size.  Physical  memory  used  by
		    process; includes shared memory. When the -F  flag
		    is	used, this value is in kilobytes, otherwise it
		    is displayed in a compact format using 4  columns.
		    In the report output, the suffixes K, M, and G are
		    decimal multipliers. That is,

		    K means x 1000, M x 1000000, and G x 1000000000.
       VSZ	    The virtual memory used by process. The format  is
		    the same as described above for RSS.
       UsrTim	    The	 user-mode  CPU	 time  being  consumed	by the
		    process.  It has two modes, depending  on  whether
		    the -ot option was specified. In the default mode,
		    the value is a normalized delta, that is, how much
		    user time has been consumed since the last sample,
		    normalized over 1 second. If  the  -ot  option  is
		    specified,	the  value  is	the absolute amount of
		    user time the process  has	accumulated  since  it
		    started, in the form Minutes:Seconds.
       SysTim	    The	 CPU time in kernel-mode being consumed by the
		    process (see the description of UsrTime above).
       Pri	    The UNIX priority of the  process.	This  is  only
		    shown when the -F option is used.
       IBk	    Input  Block Operations. Actual file system blocks
		    read or written.
       OBk	    Output Block Operations.
       Maj	    Major faults. Faults that were satisfied by	 doing
		    I/O (going to disk).
       Min	    Minor  faults.  Faults  that  were	satisfied from
		    cache.
       Command	    The name of the running program.  Arguments speci‐
		    fied   when	  the  program	was  invoked  are  not
		    retrieved.
       ────────────────────────────────────────────────────────────────

       Disks Section
       ────────────────────────────────────────────────────────────
       DSK	An index into the table that collect outputs, used
		for scripting.
       NAME	The  name of the device, specified as dskinstance,
		such as dsk23, and  found  in  the  system's  /dev
		directory.
       B/T/L	If  this  is  a SCSI disk it contains the Bus/Tar‐
		get/Lun identifier, otherwise a - (dash). Use  the
		hwmgr command to identify devices, as described in
		the hwmgr(8) reference page.
       R/S	Reads per second.
       RKB/S	Kilobytes read per second.
       W/S	Writes per second.
       WKB/S	Kilobytes written per second.
       AVS	Average service time. The time spent actually ser‐
		vicing the request -no wait time in milliseconds.
       AVW	Average	 wait  time.  The  time	 spent in the wait
		queue in milliseconds.
       ACTQ	The number of requests in the active  queue  (that
		is, being serviced by the disk).

       WTQ	The number of requests in the wait queue (have not
		yet been submitted to disk).
       %BSY	Percent Busy. The time spent servicing requests in
		interval divided by the interval.

       Tapes Section
       ────────────────────────────────────────────────────────────
       NUM	An index for scripting.
       NAME	The  device name, tape instance, where instance is
		an integer in the range 0-256 and can be found	in
		the  /dev/tape	directory.  The	 hwmgr command can
		also be used to find  devices.	See  the  hwmgr(8)
		reference  page	 for  information  on  the command
		options.
       B/T/L	The Bus/Target/Lun IDs (identifiers).
       R/S	Reads per second.
       RKB/S	Kilobytes read per second.
       W/S	Writes per second.
       WKB/S	Kilobytes written per second.

       LSM Volumes Section
       ───────────────────────────────────────────────────────────
       VOL     Index for scripting.
       NAME    Name  in	 the  form  Diskgroup/Volume  to   ensure
	       uniqueness.
       R/S     Reads per second.
       RKB/S   Kilobytes read per second.
       RAVS    Average service time for reads with respect to LSM
	       driver. (This includes disk driver wait time.)
       W/S     Writes per second.
       WKB/S   Kilobytes written per second.
       WAVS    Average service time for writes	with  respect  to
	       LSM driver. (Includes disk driver wait time.)

       CPU Summary Section
       ─────────────────────────────────────────────────────────────────
       USER...WAIT    CPU states, averaged over all CPUs.
       INTR	      Interrupts per second.
       SYSC	      System calls per second.
       CS	      Context switches per second.
       RUNQ	      Number of processes in the run queue.
       AVG5,30,60     Load average over the last 5, 30, and 60 seconds.
       FORK	      Number of forks per second.
       VFORK	      Number of vforks per second.

       Single CPU Section
       ──────────────────────────────────────────────────────────
       CPU#   Index for scripts.
       USER   Percent  time  (ticks)  spent  in user-level code.
	      This includes nice ticks.
       SYS    Percent time (ticks) spent in kernel.
       IDLE   Percent time (ticks) spent doing nothing.
       WAIT   Idle ticks while waiting for I/O to happen.

       Memory
       ──────────────────────────────────────────────────────────
       Free   Number of megabytes available.  This  is	reported
	      as pages available if you specify the -om option.
       Swap   Number  of  megabytes (or pages) available on swap
	      device(s).
       Act    Amount of active memory in megabytes (or pages).
       InAc   Amount of inactive memory in megabytes (or  pages)
	      allocated	 to a process, but marked as not used in
	      greater than X seconds.
       Wire   Nonswappable  kernel  memory  in	 megabytes   (or
	      pages).
       UBC    Megabytes (or pages) of memory used by Bufcache.
       PI     Page in operations per second.
       PO     Page out operations per second.
       Zer    Pages  zeroed  per second (overwritten with zeroes
	      before handing to a process).

       Re     Pages reactivated (status changed from inactive to
	      active).
       COW    Copies-on-write per second.
       SW     Processes swapped per second.
       HIT    UBC (unified buffer cache) hits per second.
       PP     UBC pages pushed (written to disk) per second.
       ALL    Pages allocated by UBC per second.

       Filesystem Section
       ─────────────────────────────────────────────────────────────────
       FS	     Index for scripting.
       Filesystem    Name  of file system, or the Domain#Fileset in the
		     case of an AdvFS file system. See	the  /etc/fstab
		     file  for	a  list	 of file systems present on the
		     system.
       Capacity	     In megabytes.
       Free	     In megabytes.

       Network Section
       ─────────────────────────────────────────────────────────────
       Cnt	 Index for scripting.
       Name	 Name of the network adaptor.
       Inpck	 Packets received per second.
       InErr	 Input error packets per second.
       Outpck	 Packets sent per second.
       OutErr	 Output error packets per second.
       Coll	 Collisions per second.
       IKB	 Kilobytes received per second.
       OKB	 Kilobytes sent per second.
       %BW	 Percent of theoretical bandwidth being used  (Eth‐
		 ernet = 10Mbits/sec).

       Message Queues Section
       ────────────────────────────────────────────────────────────
       ID	This is the ID according to ipcs.
       Key	The key according to ipcs.
       OUID	The  owner  UID	 (user	identifier) of the message
		queue.
       BYTES	The number of bytes in use  for	 all  messages	in
		this queue.
       Cnt	The number of messages in queue.
       SPID	The  PID  (process identifier) of the last process
		to send a message on this queue.
       RPID	The PID (process identifier) of the  last  process
		to read a message from this queue.
       STIME	The time (in epoch seconds) of the last send.
       RTIME	The time of the last receive.
       CTIME	The creation time of this queue.

       Terminal I/O Section
       ───────────────────────────────────────────────────────
       In    The number of characters input.
       Out   The number of characters output.
       Can   Portion of input characters on the CANNON queue.
       Raw   Portion of input characters on the RAW queue.

       Cluster Disks Section
       ───────────────────────────────────────────────────────────────────────
       Cnt	    An index for scripting.
       MembID	    Member ID of cluster member.
       TargetDsk    Name  of the disk device targeted by I/O from the speci‐
		    fied cluster member.
       DrdDevType   A code for the device type from the perspective  of	 DRD
		    relative	to  the	 specified  cluster member. Possible
		    codes are:

		    sd - Denotes the case when the specified member is
			 currently a server for the direct-access I/O device.

		    ss - Denotes the case when the specified member is
			 currently a server for the single-server device.

		    cs - Denotes the case when the specified member is
			 currently acting as a client for the single-server
			 device and it can potentially function as a server
			 for the device.

		    co - Denotes the case when the specified member is
			 currently acting as a client for the (direct-access
			 I/O or single-server) device and it is not a
			 potential server for the device.

		    un - Denotes the case when it cannot be determined
			 whether the  device is of type sd, ss, cs or co
			 relative to the specified member.

       R/S	    Reads per second.
       RKB/S	    Kilobytes read per second.
       W/S	    Writes per second.
       WKB/S	    Kilobytes written per second.

       Cluster Tapes Section
       ─────────────────────────────────────────────────────────────────────
       Cnt	    An index for scripting.
       MembID	    Member ID of cluster member.
       TargetTape   Name of the tape device targeted by I/O from the speci‐
		    fied cluster member.
       DrdDevType   A  code for the device type from the perspective of DRD
		    relative to	 the  specified	 cluster  member.  Possible
		    codes  for	cluster tape devices are ss, cs, co and un.
		    For descriptions of these codes, refer to  the  Cluster
		    Disks Section.
       R/S	    Reads per second.
       RKB/S	    Kilobytes read per second.
       W/S	    Writes per second.
       WKB/S	    Kilobytes written per second.

RESTRICTIONS
       The  following  restrictions apply when using collect: The average ser‐
       vice time for storage units made available by the MYLEX	(SWXCR)	 host-
       based  hardware	RAI  controller is not available.  The collect utility
       cannot dynamically recognize new devices or hardware added to the  sys‐
       tem while collect is running. If you run collect and then install a new
       disk or tape device, and start using that device, collect cannot gather
       data on the newly-installed device. The same is true of any LSM volumes
       created on newly-installed disks. There is one exception: collect (data
       file  version  15  and above) will recognize the addition or removal of
       CPUs.

	      To resolve this problem, restart collect after adding new	 hard‐
	      ware to the system.  Statistics for ISDN PPP connections are not
	      available.

EXAMPLES
       The following example shows how to run a full data collection and  dis‐
       play  the output at the terminal using the standard interval of 10 sec‐
       onds: # collect

	      This command is similar to the output monitoring	commands  such
	      as  vmstat, iostat, netstat, volstat, ipcs, and ps.  The follow‐
	      ing command uses the -s option to collect only process  informa‐
	      tion  in	the  file  sys.data.  The -S option specifies that the
	      data is sorted by CPU usage, and the -n  option  specifies  that
	      the top ten processes are saved:
	       # collect -sp -S -n10 -f sys.data Initializing (10.0 seconds)

	      The message Initializing (10.0 seconds) indicates that data col‐
	      lection will be performed at 10-second intervals.	 The following
	      command  displays the data collected in the preceding example by
	      piping the output to the more command: # collect -p sys.data.cgz
	      |								  more
	      ###############################################################
	      OSF1   glop.ytx.tog.com	T5.0   77.11   DEC1000	1/266MHz/256MB
	      HOST............glop.ytx.tog.cm  Started.<DY:MM:DT:HH:MM:SS:YR>
					       Seconds........943298217

	       CPU FAMILY......21064 (EV4 core) CPU ID.........EV4.5 (21064)
	       CPU EXTENSIONS..
	       PLATFORM NAME...DEC1000		CPU SPEED......266 MHz
	       SWAP SIZE.......196 MB		Physical Mem...256 MB
	       NUM CPUS........1		NUM DISKS......3
	       NUM LANS........3		NUM FSYS.......4
	       MAX MQUEUES.....64		NUM TAPES......0
	       INTERVAL........10.00		PROC_INTERVAL..10.00
	       UBCMAXPERCENT...100		UBCMINPERCENT..10
	       MAXUSERS........256		MAXUPRC........64
	       Delay_WBuffers..0		       LSM	  Volumes....0
	      ###############################################################

	      #### RECORD    1 (943298227:10) (Mon Nov 22 14:17:07 2000) ####

	      Process Statistics (RSS & VSZ in KBytes) PID  User %CPU RSS  VSZ
	      UsrTim SysTim IBk OBk Maj Min   Command
		 0 root 1.7  12M  342M	0.00  0.00  0	0   0	0 kernel  idle
	      3275 root 0.3  3.3M 5.6M	0.00  0.00  0	0   0	8     collect
	       482 root 0.0  2.6M 6.3M	0.00  0.00  0	0   0	0    insightd
	       360  root 0.0  2.0M 4.4M	 0.00  0.00  0	 0   0	 0   automount
	      .	 .  .

	      Note that the preceding sample report is modified and compressed
	      for ease of reference. It might appear wider on your terminal or
	      in a printed report.  The following command uses the  -e	option
	      to  exclude  file	 system	 data  and collects data every second,
	      except for process data, which is collected every 5 seconds. The
	      times  are  set  using  the  -i  option.	# collect -ef -i1,5 -f
	      sys.data Initializing (1.0 seconds) ... done.

	      Note that the time has changed in	 the  initialization  message.
	      The  following  command prints only the header section of a data
	      file. That is the information bordered by the  hash  (or	pound)
	      symbol,  (#)  as shown in the sample output in Example 3: # col‐
	      lect -sh -p sys.data #### RECORD	 1 (943298227:10) (Mon Nov  22
	      14:17:07	2000) #### The following command selects only the data
	      from the network	subsystem  and	displays  it  at  the  command
	      prompt: # collect -sn Initializing (10.0 seconds) ... done.

	      ### RECORD 1 (943045470:0) (Fri Nov 19 16:04:30 2000) ###

	      #	  Network  Statistics  #Cnt	Name Inpck InErr Outpck OutErr
	      Coll  IKB	 OKB %BW
		 0     lo0     0     0	    0	   0	0    0	  0   0
		 1     sl0     0     0	    0	   0	0    0	  0   0
		 2     tu0    75     0	    0	   0	0    8	   0	0  The
	      following	 command  specifies only data from the disk subsystem,
	      and then only from specific disks identified as dsk0, dsk1,  and
	      dsk8. The disk names are determined by their device special file
	      names in the /dev/disk directory.	 The disk names are entered on
	      the  command  line separated by commas, with no blank spaces, as
	      shown in this example: # collect -sd -Ddsk0,dsk1,dsk8 Initializ‐
	      ing (1.0 seconds) ... done.

	      The  hwmgr  command  can	also be used to find devices.  See the
	      hwmgr(8) reference page for information on the command  options.
	      The  following command shows how to use the -p option to convert
	      data files created using a previous version of the collect util‐
	      ity:  #  collect	-p /tmp/olddata.col -f \ /tmp/oldconverted.col
	      Initializing (1.0 seconds) ... done.

FILES
       The executable image.

SEE ALSO
       Commands: sys_check(8), hwmgr(8), drdmgr(8)

       Manuals: System Configuration and Tuning and System Administration

								    collect(8)
[top]
                             _         _         _ 
                            | |       | |       | |     
                            | |       | |       | |     
                         __ | | __ __ | | __ __ | | __  
                         \ \| |/ / \ \| |/ / \ \| |/ /  
                          \ \ / /   \ \ / /   \ \ / /   
                           \   /     \   /     \   /    
                            \_/       \_/       \_/ 
More information is available in HTML format for server OSF1

List of man pages available for OSF1

Copyright (c) for man pages and the logo by the respective OS vendor.

For those who want to learn more, the polarhome community provides shell access and support.

[legal] [privacy] [GNU] [policy] [cookies] [netiquette] [sponsors] [FAQ]
Tweet
Polarhome, production since 1999.
Member of Polarhome portal.
Based on Fawad Halim's script.
....................................................................
Vote for polarhome
Free Shell Accounts :: the biggest list on the net