1SAR(1) Linux User's Manual SAR(1)
2
6 sar - Collect, report, or save system activity information.
7
9 sar [ -A ] [ -B ] [ -b ] [ -C ] [ -D ] [ -d ] [ -F [ MOUNT ] ] [ -H ] [
10 -h ] [ -p ] [ -q ] [ -r [ ALL ] ] [ -S ] [ -t ] [ -u [ ALL ] ] [ -V ] [
11 -v ] [ -W ] [ -w ] [ -y ] [ -z ] [ --help ] [ --human ] [ --sadc ] [ -I
12 { int_list | SUM | ALL } ] [ -P { cpu_list | ALL } ] [ -m { keyword
13 [,...] | ALL } ] [ -n { keyword [,...] | ALL } ] [ -j { ID | LABEL |
14 PATH | UUID | ... } ] [ -f [ filename ] | -o [ filename ] | -[0-9]+ ] [
15 -i interval ] [ -s [ hh:mm[:ss] ] ] [ -e [ hh:mm[:ss] ] ] [ interval [
16 count ] ]
17
19 The sar command writes to standard output the contents of selected
20 cumulative activity counters in the operating system. The accounting
21 system, based on the values in the count and interval parameters,
22 writes information the specified number of times spaced at the speci‐
23 fied intervals in seconds. If the interval parameter is set to zero,
24 the sar command displays the average statistics for the time since the
25 system was started. If the interval parameter is specified without the
26 count parameter, then reports are generated continuously. The col‐
27 lected data can also be saved in the file specified by the -o filename
28 flag, in addition to being displayed onto the screen. If filename is
29 omitted, sar uses the standard system activity daily data file (see
30 below). By default all the data available from the kernel are saved in
31 the data file.
32 The sar command extracts and writes to standard output records previ‐
34 ously saved in a file. This file can be either the one specified by the
35 -f flag or, by default, the standard system activity daily data file.
36 It is also possible to enter -1, -2 etc. as an argument to sar to dis‐
37 play data of that days ago. For example, -1 will point at the standard
38 system activity file of yesterday.
39 Standard system activity daily data files are named saDD or saYYYYMMDD,
41 where YYYY stands for the current year, MM for the current month and DD
42 for the current day. They are the default files used by sar only when
43 no filename has been explicitly specified. When used to write data to
44 files (with its option -o), sar will use saYYYYMMDD if option -D has
45 also been specified, else it will use saDD. When used to display the
46 records previously saved in a file, sar will look for the most recent
47 of saDD and saYYYYMMDD, and use it.
48 Standard system activity daily data files are located in the
50 /var/log/sa directory by default. Yet it is possible to specify an
51 alternate location for them: If a directory (instead of a plain file)
52 is used with options -f or -o then it will be considered as the direc‐
53 tory containing the data files.
54 Without the -P flag, the sar command reports system-wide (global among
56 all processors) statistics, which are calculated as averages for values
57 expressed as percentages, and as sums otherwise. If the -P flag is
58 given, the sar command reports activity which relates to the specified
59 processor or processors. If -P ALL is given, the sar command reports
60 statistics for each individual processor and global statistics among
61 all processors. Offline processors are not displayed.
62 You can select information about specific system activities using
64 flags. Not specifying any flags selects only CPU activity. Specifying
65 the -A flag selects all possible activities.
66 The default version of the sar command (CPU utilization report) might
68 be one of the first facilities the user runs to begin system activity
69 investigation, because it monitors major system resources. If CPU uti‐
70 lization is near 100 percent (user + nice + system), the workload sam‐
71 pled is CPU-bound.
72 If multiple samples and multiple reports are desired, it is convenient
74 to specify an output file for the sar command. Run the sar command as
75 a background process. The syntax for this is:
76 sar -o datafile interval count >/dev/null 2>&1 &
78 All data are captured in binary form and saved to a file (datafile).
80 The data can then be selectively displayed with the sar command using
81 the -f option. Set the interval and count parameters to select count
82 records at interval second intervals. If the count parameter is not
83 set, all the records saved in the file will be selected. Collection of
84 data in this manner is useful to characterize system usage over a
85 period of time and determine peak usage hours.
86 Note: The sar command only reports on local activities.
88
91 -A This is equivalent to specifying -bBdFHqSuvwWy -I SUM -I ALL -m
92 ALL -n ALL -r ALL -u ALL -P ALL.
93 -B Report paging statistics. The following values are displayed:
95 pgpgin/s
97 Total number of kilobytes the system paged in from disk
98 per second.
99 pgpgout/s
101 Total number of kilobytes the system paged out to disk
102 per second.
103 fault/s
105 Number of page faults (major + minor) made by the system
106 per second. This is not a count of page faults that gen‐
107 erate I/O, because some page faults can be resolved with‐
108 out I/O.
109 majflt/s
111 Number of major faults the system has made per second,
112 those which have required loading a memory page from
113 disk.
114 pgfree/s
116 Number of pages placed on the free list by the system per
117 second.
118 pgscank/s
120 Number of pages scanned by the kswapd daemon per second.
121 pgscand/s
123 Number of pages scanned directly per second.
124 pgsteal/s
126 Number of pages the system has reclaimed from cache
127 (pagecache and swapcache) per second to satisfy its mem‐
128 ory demands.
129 %vmeff
131 Calculated as pgsteal / pgscan, this is a metric of the
132 efficiency of page reclaim. If it is near 100% then
133 almost every page coming off the tail of the inactive
134 list is being reaped. If it gets too low (e.g. less than
135 30%) then the virtual memory is having some difficulty.
136 This field is displayed as zero if no pages have been
137 scanned during the interval of time.
138 -b Report I/O and transfer rate statistics. The following values
140 are displayed:
141 tps
143 Total number of transfers per second that were issued to
144 physical devices. A transfer is an I/O request to a
145 physical device. Multiple logical requests can be com‐
146 bined into a single I/O request to the device. A trans‐
147 fer is of indeterminate size.
148 rtps
150 Total number of read requests per second issued to physi‐
151 cal devices.
152 wtps
154 Total number of write requests per second issued to phys‐
155 ical devices.
156 bread/s
158 Total amount of data read from the devices in blocks per
159 second. Blocks are equivalent to sectors and therefore
160 have a size of 512 bytes.
161 bwrtn/s
163 Total amount of data written to devices in blocks per
164 second.
165 -C When reading data from a file, tell sar to display comments that
167 have been inserted by sadc.
168 -D Use saYYYYMMDD instead of saDD as the standard system activity
170 daily data file name. This option works only when used in con‐
171 junction with option -o to save data to file.
172 -d Report activity for each block device. When data are displayed,
174 the device specification devM-n is generally used (DEV column).
175 M is the major number of the device and n its minor number.
176 Device names may also be pretty-printed if option -p is used or
177 persistent device names can be printed if option -j is used (see
178 below). Note that disk activity depends on sadc options -S DISK
179 and -S XDISK to be collected. The following values are dis‐
180 played:
181 tps
183 Total number of transfers per second that were issued to
184 physical devices. A transfer is an I/O request to a
185 physical device. Multiple logical requests can be com‐
186 bined into a single I/O request to the device. A trans‐
187 fer is of indeterminate size.
188 rkB/s
190 Number of kilobytes read from the device per second.
191 wkB/s
193 Number of kilobytes written to the device per second.
194 areq-sz
196 The average size (in kilobytes) of the I/O requests that
197 were issued to the device.
198 Note: In previous versions, this field was known as
199 avgrq-sz and was expressed in sectors.
200 aqu-sz
202 The average queue length of the requests that were issued
203 to the device.
204 Note: In previous versions, this field was known as
205 avgqu-sz.
206 await
208 The average time (in milliseconds) for I/O requests
209 issued to the device to be served. This includes the time
210 spent by the requests in queue and the time spent servic‐
211 ing them.
212 svctm
214 The average service time (in milliseconds) for I/O
215 requests that were issued to the device. Warning! Do not
216 trust this field any more. This field will be removed in
217 a future sysstat version.
218 %util
220 Percentage of elapsed time during which I/O requests were
221 issued to the device (bandwidth utilization for the
222 device). Device saturation occurs when this value is
223 close to 100% for devices serving requests serially. But
224 for devices serving requests in parallel, such as RAID
225 arrays and modern SSDs, this number does not reflect
226 their performance limits.
227 -e [ hh:mm[:ss] ]
229 Set the ending time of the report. The default ending time is
230 18:00:00. Hours must be given in 24-hour format. This option
231 can be used when data are read from or written to a file
232 (options -f or -o).
233 -F [ MOUNT ]
235 Display statistics for currently mounted filesystems. Pseudo-
236 filesystems are ignored. At the end of the report, sar will dis‐
237 play a summary of all those filesystems. Use of the MOUNT
238 parameter keyword indicates that mountpoint will be reported
239 instead of filesystem device. Note that filesystems statistics
240 depend on sadc option -S XDISK to be collected.
241 The following values are displayed:
243 MBfsfree
245 Total amount a free space in megabytes (including space
246 available only to privileged user).
247 MBfsused
249 Total amount of space used in megabytes.
250 %fsused
252 Percentage of filesystem space used, as seen by a privi‐
253 leged user.
254 %ufsused
256 Percentage of filesystem space used, as seen by an
257 unprivileged user.
258 Ifree
260 Total number of free file nodes in filesystem.
261 Iused
263 Total number of file nodes used in filesystem.
264 %Iused
266 Percentage of file nodes used in filesystem.
267 -f [ filename ]
269 Extract records from filename (created by the -o filename flag).
270 The default value of the filename parameter is the current stan‐
271 dard system activity daily data file. If filename is a direc‐
272 tory instead of a plain file then it is considered as the direc‐
273 tory where the standard system activity daily data files are
274 located. The -f option is exclusive of the -o option.
275 -H Report hugepages utilization statistics. The following values
277 are displayed:
278 kbhugfree
280 Amount of hugepages memory in kilobytes that is not yet
281 allocated.
282 kbhugused
284 Amount of hugepages memory in kilobytes that has been
285 allocated.
286 %hugused
288 Percentage of total hugepages memory that has been allo‐
289 cated.
290 -h Make the output of sar easier to read by a human. Options
292 --human and -p (pretty-print) are enabled implicitly with this
293 option.
294 --help Display a short help message then exit.
296 --human
298 Print sizes in human readable format (e.g. 1.0k, 1.2M, etc.)
299 The units displayed with this option supersede any other default
300 units (e.g. kilobytes, sectors...) associated with the metrics.
301 -I { int_list | SUM | ALL }
303 Report statistics for interrupts. int_list is a list of comma-
304 separated values or range of values (e.g., 0-16,35,400-). The
305 SUM keyword indicates that the total number of interrupts
306 received per second is to be displayed. The ALL keyword indi‐
307 cates that statistics from all interrupts, including potential
308 APIC interrupt sources, are to be reported. Note that interrupt
309 statistics depend on sadc option "-S INT" to be collected.
310 -i interval
312 Select data records at seconds as close as possible to the num‐
313 ber specified by the interval parameter.
314 -j { ID | LABEL | PATH | UUID | ... }
316 Display persistent device names. Use this option in conjunction
317 with option -d. Options ID, LABEL, etc. specify the type of the
318 persistent name. These options are not limited, only prerequi‐
319 site is that directory with required persistent names is present
320 in /dev/disk. If persistent name is not found for the device,
321 the device name is pretty-printed (see option -p below).
322 -m { keyword [,...] | ALL }
324 Report power management statistics. Note that these statistics
325 depend on sadc's option "-S POWER" to be collected.
326 Possible keywords are CPU, FAN, FREQ, IN, TEMP and USB.
328 With the CPU keyword, statistics about CPU are reported. The
330 following value is displayed:
331 MHz
333 Instantaneous CPU clock frequency in MHz.
334 With the FAN keyword, statistics about fans speed are reported.
336 The following values are displayed:
337 rpm
339 Fan speed expressed in revolutions per minute.
340 drpm
342 This field is calculated as the difference between cur‐
343 rent fan speed (rpm) and its low limit (fan_min).
344 DEVICE
346 Sensor device name.
347 With the FREQ keyword, statistics about CPU clock frequency are
349 reported. The following value is displayed:
350 wghMHz
352 Weighted average CPU clock frequency in MHz. Note that
353 the cpufreq-stats driver must be compiled in the kernel
354 for this option to work.
355 With the IN keyword, statistics about voltage inputs are
357 reported. The following values are displayed:
358 inV
360 Voltage input expressed in Volts.
361 %in
363 Relative input value. A value of 100% means that voltage
364 input has reached its high limit (in_max) whereas a value
365 of 0% means that it has reached its low limit (in_min).
366 DEVICE
368 Sensor device name.
369 With the TEMP keyword, statistics about devices temperature are
371 reported. The following values are displayed:
372 degC
374 Device temperature expressed in degrees Celsius.
375 %temp
377 Relative device temperature. A value of 100% means that
378 temperature has reached its high limit (temp_max).
379 DEVICE
381 Sensor device name.
382 With the USB keyword, the sar command takes a snapshot of all
384 the USB devices currently plugged into the system. At the end of
385 the report, sar will display a summary of all those USB devices.
386 The following values are displayed:
387 BUS
389 Root hub number of the USB device.
390 idvendor
392 Vendor ID number (assigned by USB organization).
393 idprod
395 Product ID number (assigned by Manufacturer).
396 maxpower
398 Maximum power consumption of the device (expressed in
399 mA).
400 manufact
402 Manufacturer name.
403 product
405 Product name.
406 The ALL keyword is equivalent to specifying all the keywords
408 above and therefore all the power management statistics are
409 reported.
410 -n { keyword [,...] | ALL }
412 Report network statistics.
413 Possible keywords are DEV, EDEV, FC, ICMP, EICMP, ICMP6, EICMP6,
415 IP, EIP, IP6, EIP6, NFS, NFSD, SOCK, SOCK6, SOFT, TCP, ETCP, UDP
416 and UDP6.
417 With the DEV keyword, statistics from the network devices are
419 reported. The following values are displayed:
420 IFACE
422 Name of the network interface for which statistics are
423 reported.
424 rxpck/s
426 Total number of packets received per second.
427 txpck/s
429 Total number of packets transmitted per second.
430 rxkB/s
432 Total number of kilobytes received per second.
433 txkB/s
435 Total number of kilobytes transmitted per second.
436 rxcmp/s
438 Number of compressed packets received per second (for
439 cslip etc.).
440 txcmp/s
442 Number of compressed packets transmitted per second.
443 rxmcst/s
445 Number of multicast packets received per second.
446 %ifutil
448 Utilization percentage of the network interface. For
449 half-duplex interfaces, utilization is calculated using
450 the sum of rxkB/s and txkB/s as a percentage of the
451 interface speed. For full-duplex, this is the greater of
452 rxkB/S or txkB/s.
453 With the EDEV keyword, statistics on failures (errors) from the
455 network devices are reported. The following values are dis‐
456 played:
457 IFACE
459 Name of the network interface for which statistics are
460 reported.
461 rxerr/s
463 Total number of bad packets received per second.
464 txerr/s
466 Total number of errors that happened per second while
467 transmitting packets.
468 coll/s
470 Number of collisions that happened per second while
471 transmitting packets.
472 rxdrop/s
474 Number of received packets dropped per second because of
475 a lack of space in linux buffers.
476 txdrop/s
478 Number of transmitted packets dropped per second because
479 of a lack of space in linux buffers.
480 txcarr/s
482 Number of carrier-errors that happened per second while
483 transmitting packets.
484 rxfram/s
486 Number of frame alignment errors that happened per second
487 on received packets.
488 rxfifo/s
490 Number of FIFO overrun errors that happened per second on
491 received packets.
492 txfifo/s
494 Number of FIFO overrun errors that happened per second on
495 transmitted packets.
496 With the FC keyword, statistics about fibre channel traffic are
498 reported. Note that fibre channel statistics depend on sadc's
499 option "-S DISK" to be collected. The following values are dis‐
500 played:
501 FCHOST
503 Name of the fibre channel host bus adapter (HBA) inter‐
504 face for which statistics are reported.
505 fch_rxf/s
507 The total number of frames received per second.
508 fch_txf/s
510 The total number of frames transmitted per second.
511 fch_rxw/s
513 The total number of transmission words received per sec‐
514 ond.
515 fch_txw/s
517 The total number of transmission words transmitted per
518 second.
519 With the ICMP keyword, statistics about ICMPv4 network traffic
521 are reported. Note that ICMPv4 statistics depend on sadc's
522 option "-S SNMP" to be collected. The following values are dis‐
523 played (formal SNMP names between square brackets):
524 imsg/s
526 The total number of ICMP messages which the entity
527 received per second [icmpInMsgs]. Note that this counter
528 includes all those counted by ierr/s.
529 omsg/s
531 The total number of ICMP messages which this entity
532 attempted to send per second [icmpOutMsgs]. Note that
533 this counter includes all those counted by oerr/s.
534 iech/s
536 The number of ICMP Echo (request) messages received per
537 second [icmpInEchos].
538 iechr/s
540 The number of ICMP Echo Reply messages received per sec‐
541 ond [icmpInEchoReps].
542 oech/s
544 The number of ICMP Echo (request) messages sent per sec‐
545 ond [icmpOutEchos].
546 oechr/s
548 The number of ICMP Echo Reply messages sent per second
549 [icmpOutEchoReps].
550 itm/s
552 The number of ICMP Timestamp (request) messages received
553 per second [icmpInTimestamps].
554 itmr/s
556 The number of ICMP Timestamp Reply messages received per
557 second [icmpInTimestampReps].
558 otm/s
560 The number of ICMP Timestamp (request) messages sent per
561 second [icmpOutTimestamps].
562 otmr/s
564 The number of ICMP Timestamp Reply messages sent per sec‐
565 ond [icmpOutTimestampReps].
566 iadrmk/s
568 The number of ICMP Address Mask Request messages received
569 per second [icmpInAddrMasks].
570 iadrmkr/s
572 The number of ICMP Address Mask Reply messages received
573 per second [icmpInAddrMaskReps].
574 oadrmk/s
576 The number of ICMP Address Mask Request messages sent per
577 second [icmpOutAddrMasks].
578 oadrmkr/s
580 The number of ICMP Address Mask Reply messages sent per
581 second [icmpOutAddrMaskReps].
582 With the EICMP keyword, statistics about ICMPv4 error messages
584 are reported. Note that ICMPv4 statistics depend on sadc's
585 option "-S SNMP" to be collected. The following values are dis‐
586 played (formal SNMP names between square brackets):
587 ierr/s
589 The number of ICMP messages per second which the entity
590 received but determined as having ICMP-specific errors
591 (bad ICMP checksums, bad length, etc.) [icmpInErrors].
592 oerr/s
594 The number of ICMP messages per second which this entity
595 did not send due to problems discovered within ICMP such
596 as a lack of buffers [icmpOutErrors].
597 idstunr/s
599 The number of ICMP Destination Unreachable messages
600 received per second [icmpInDestUnreachs].
601 odstunr/s
603 The number of ICMP Destination Unreachable messages sent
604 per second [icmpOutDestUnreachs].
605 itmex/s
607 The number of ICMP Time Exceeded messages received per
608 second [icmpInTimeExcds].
609 otmex/s
611 The number of ICMP Time Exceeded messages sent per second
612 [icmpOutTimeExcds].
613 iparmpb/s
615 The number of ICMP Parameter Problem messages received
616 per second [icmpInParmProbs].
617 oparmpb/s
619 The number of ICMP Parameter Problem messages sent per
620 second [icmpOutParmProbs].
621 isrcq/s
623 The number of ICMP Source Quench messages received per
624 second [icmpInSrcQuenchs].
625 osrcq/s
627 The number of ICMP Source Quench messages sent per second
628 [icmpOutSrcQuenchs].
629 iredir/s
631 The number of ICMP Redirect messages received per second
632 [icmpInRedirects].
633 oredir/s
635 The number of ICMP Redirect messages sent per second
636 [icmpOutRedirects].
637 With the ICMP6 keyword, statistics about ICMPv6 network traffic
639 are reported. Note that ICMPv6 statistics depend on sadc's
640 option "-S IPV6" to be collected. The following values are dis‐
641 played (formal SNMP names between square brackets):
642 imsg6/s
644 The total number of ICMP messages received by the inter‐
645 face per second which includes all those counted by
646 ierr6/s [ipv6IfIcmpInMsgs].
647 omsg6/s
649 The total number of ICMP messages which this interface
650 attempted to send per second [ipv6IfIcmpOutMsgs].
651 iech6/s
653 The number of ICMP Echo (request) messages received by
654 the interface per second [ipv6IfIcmpInEchos].
655 iechr6/s
657 The number of ICMP Echo Reply messages received by the
658 interface per second [ipv6IfIcmpInEchoReplies].
659 oechr6/s
661 The number of ICMP Echo Reply messages sent by the inter‐
662 face per second [ipv6IfIcmpOutEchoReplies].
663 igmbq6/s
665 The number of ICMPv6 Group Membership Query messages
666 received by the interface per second [ipv6IfIcmpInGroup‐
667 MembQueries].
668 igmbr6/s
670 The number of ICMPv6 Group Membership Response messages
671 received by the interface per second [ipv6IfIcmpInGroup‐
672 MembResponses].
673 ogmbr6/s
675 The number of ICMPv6 Group Membership Response messages
676 sent per second [ipv6IfIcmpOutGroupMembResponses].
677 igmbrd6/s
679 The number of ICMPv6 Group Membership Reduction messages
680 received by the interface per second [ipv6IfIcmpInGroup‐
681 MembReductions].
682 ogmbrd6/s
684 The number of ICMPv6 Group Membership Reduction messages
685 sent per second [ipv6IfIcmpOutGroupMembReductions].
686 irtsol6/s
688 The number of ICMP Router Solicit messages received by
689 the interface per second [ipv6IfIcmpInRouterSolicits].
690 ortsol6/s
692 The number of ICMP Router Solicitation messages sent by
693 the interface per second [ipv6IfIcmpOutRouterSolicits].
694 irtad6/s
696 The number of ICMP Router Advertisement messages received
697 by the interface per second [ipv6IfIcmpInRouterAdvertise‐
698 ments].
699 inbsol6/s
701 The number of ICMP Neighbor Solicit messages received by
702 the interface per second [ipv6IfIcmpInNeighborSolicits].
703 onbsol6/s
705 The number of ICMP Neighbor Solicitation messages sent by
706 the interface per second [ipv6IfIcmpOutNeighborSolicits].
707 inbad6/s
709 The number of ICMP Neighbor Advertisement messages
710 received by the interface per second [ipv6IfIcmpInNeighb‐
711 orAdvertisements].
712 onbad6/s
714 The number of ICMP Neighbor Advertisement messages sent
715 by the interface per second [ipv6IfIcmpOutNeighborAdver‐
716 tisements].
717 With the EICMP6 keyword, statistics about ICMPv6 error messages
719 are reported. Note that ICMPv6 statistics depend on sadc's
720 option "-S IPV6" to be collected. The following values are dis‐
721 played (formal SNMP names between square brackets):
722 ierr6/s
724 The number of ICMP messages per second which the inter‐
725 face received but determined as having ICMP-specific
726 errors (bad ICMP checksums, bad length, etc.)
727 [ipv6IfIcmpInErrors]
728 idtunr6/s
730 The number of ICMP Destination Unreachable messages
731 received by the interface per second [ipv6IfIcmpInDestUn‐
732 reachs].
733 odtunr6/s
735 The number of ICMP Destination Unreachable messages sent
736 by the interface per second [ipv6IfIcmpOutDestUnreachs].
737 itmex6/s
739 The number of ICMP Time Exceeded messages received by the
740 interface per second [ipv6IfIcmpInTimeExcds].
741 otmex6/s
743 The number of ICMP Time Exceeded messages sent by the
744 interface per second [ipv6IfIcmpOutTimeExcds].
745 iprmpb6/s
747 The number of ICMP Parameter Problem messages received by
748 the interface per second [ipv6IfIcmpInParmProblems].
749 oprmpb6/s
751 The number of ICMP Parameter Problem messages sent by the
752 interface per second [ipv6IfIcmpOutParmProblems].
753 iredir6/s
755 The number of Redirect messages received by the interface
756 per second [ipv6IfIcmpInRedirects].
757 oredir6/s
759 The number of Redirect messages sent by the interface by
760 second [ipv6IfIcmpOutRedirects].
761 ipck2b6/s
763 The number of ICMP Packet Too Big messages received by
764 the interface per second [ipv6IfIcmpInPktTooBigs].
765 opck2b6/s
767 The number of ICMP Packet Too Big messages sent by the
768 interface per second [ipv6IfIcmpOutPktTooBigs].
769 With the IP keyword, statistics about IPv4 network traffic are
771 reported. Note that IPv4 statistics depend on sadc's option "-S
772 SNMP" to be collected. The following values are displayed (for‐
773 mal SNMP names between square brackets):
774 irec/s
776 The total number of input datagrams received from inter‐
777 faces per second, including those received in error
778 [ipInReceives].
779 fwddgm/s
781 The number of input datagrams per second, for which this
782 entity was not their final IP destination, as a result of
783 which an attempt was made to find a route to forward them
784 to that final destination [ipForwDatagrams].
785 idel/s
787 The total number of input datagrams successfully deliv‐
788 ered per second to IP user-protocols (including ICMP)
789 [ipInDelivers].
790 orq/s
792 The total number of IP datagrams which local IP user-pro‐
793 tocols (including ICMP) supplied per second to IP in
794 requests for transmission [ipOutRequests]. Note that
795 this counter does not include any datagrams counted in
796 fwddgm/s.
797 asmrq/s
799 The number of IP fragments received per second which
800 needed to be reassembled at this entity [ipReasmReqds].
801 asmok/s
803 The number of IP datagrams successfully re-assembled per
804 second [ipReasmOKs].
805 fragok/s
807 The number of IP datagrams that have been successfully
808 fragmented at this entity per second [ipFragOKs].
809 fragcrt/s
811 The number of IP datagram fragments that have been gener‐
812 ated per second as a result of fragmentation at this
813 entity [ipFragCreates].
814 With the EIP keyword, statistics about IPv4 network errors are
816 reported. Note that IPv4 statistics depend on sadc's option "-S
817 SNMP" to be collected. The following values are displayed (for‐
818 mal SNMP names between square brackets):
819 ihdrerr/s
821 The number of input datagrams discarded per second due to
822 errors in their IP headers, including bad checksums, ver‐
823 sion number mismatch, other format errors, time-to-live
824 exceeded, errors discovered in processing their IP
825 options, etc. [ipInHdrErrors]
826 iadrerr/s
828 The number of input datagrams discarded per second
829 because the IP address in their IP header's destination
830 field was not a valid address to be received at this
831 entity. This count includes invalid addresses (e.g.,
832 0.0.0.0) and addresses of unsupported Classes (e.g.,
833 Class E). For entities which are not IP routers and
834 therefore do not forward datagrams, this counter includes
835 datagrams discarded because the destination address was
836 not a local address [ipInAddrErrors].
837 iukwnpr/s
839 The number of locally-addressed datagrams received suc‐
840 cessfully but discarded per second because of an unknown
841 or unsupported protocol [ipInUnknownProtos].
842 idisc/s
844 The number of input IP datagrams per second for which no
845 problems were encountered to prevent their continued pro‐
846 cessing, but which were discarded (e.g., for lack of buf‐
847 fer space) [ipInDiscards]. Note that this counter does
848 not include any datagrams discarded while awaiting re-
849 assembly.
850 odisc/s
852 The number of output IP datagrams per second for which no
853 problem was encountered to prevent their transmission to
854 their destination, but which were discarded (e.g., for
855 lack of buffer space) [ipOutDiscards]. Note that this
856 counter would include datagrams counted in fwddgm/s if
857 any such packets met this (discretionary) discard crite‐
858 rion.
859 onort/s
861 The number of IP datagrams discarded per second because
862 no route could be found to transmit them to their desti‐
863 nation [ipOutNoRoutes]. Note that this counter includes
864 any packets counted in fwddgm/s which meet this 'no-
865 route' criterion. Note that this includes any datagrams
866 which a host cannot route because all of its default
867 routers are down.
868 asmf/s
870 The number of failures detected per second by the IP re-
871 assembly algorithm (for whatever reason: timed out,
872 errors, etc) [ipReasmFails]. Note that this is not nec‐
873 essarily a count of discarded IP fragments since some
874 algorithms can lose track of the number of fragments by
875 combining them as they are received.
876 fragf/s
878 The number of IP datagrams that have been discarded per
879 second because they needed to be fragmented at this
880 entity but could not be, e.g., because their Don't Frag‐
881 ment flag was set [ipFragFails].
882 With the IP6 keyword, statistics about IPv6 network traffic are
884 reported. Note that IPv6 statistics depend on sadc's option "-S
885 IPV6" to be collected. The following values are displayed (for‐
886 mal SNMP names between square brackets):
887 irec6/s
889 The total number of input datagrams received from inter‐
890 faces per second, including those received in error
891 [ipv6IfStatsInReceives].
892 fwddgm6/s
894 The number of output datagrams per second which this
895 entity received and forwarded to their final destinations
896 [ipv6IfStatsOutForwDatagrams].
897 idel6/s
899 The total number of datagrams successfully delivered per
900 second to IPv6 user-protocols (including ICMP) [ipv6IfS‐
901 tatsInDelivers].
902 orq6/s
904 The total number of IPv6 datagrams which local IPv6 user-
905 protocols (including ICMP) supplied per second to IPv6 in
906 requests for transmission [ipv6IfStatsOutRequests]. Note
907 that this counter does not include any datagrams counted
908 in fwddgm6/s.
909 asmrq6/s
911 The number of IPv6 fragments received per second which
912 needed to be reassembled at this interface [ipv6IfStat‐
913 sReasmReqds].
914 asmok6/s
916 The number of IPv6 datagrams successfully reassembled per
917 second [ipv6IfStatsReasmOKs].
918 imcpck6/s
920 The number of multicast packets received per second by
921 the interface [ipv6IfStatsInMcastPkts].
922 omcpck6/s
924 The number of multicast packets transmitted per second by
925 the interface [ipv6IfStatsOutMcastPkts].
926 fragok6/s
928 The number of IPv6 datagrams that have been successfully
929 fragmented at this output interface per second [ipv6IfS‐
930 tatsOutFragOKs].
931 fragcr6/s
933 The number of output datagram fragments that have been
934 generated per second as a result of fragmentation at this
935 output interface [ipv6IfStatsOutFragCreates].
936 With the EIP6 keyword, statistics about IPv6 network errors are
938 reported. Note that IPv6 statistics depend on sadc's option "-S
939 IPV6" to be collected. The following values are displayed (for‐
940 mal SNMP names between square brackets):
941 ihdrer6/s
943 The number of input datagrams discarded per second due to
944 errors in their IPv6 headers, including version number
945 mismatch, other format errors, hop count exceeded, errors
946 discovered in processing their IPv6 options, etc.
947 [ipv6IfStatsInHdrErrors]
948 iadrer6/s
950 The number of input datagrams discarded per second
951 because the IPv6 address in their IPv6 header's destina‐
952 tion field was not a valid address to be received at this
953 entity. This count includes invalid addresses (e.g., ::0)
954 and unsupported addresses (e.g., addresses with unallo‐
955 cated prefixes). For entities which are not IPv6 routers
956 and therefore do not forward datagrams, this counter
957 includes datagrams discarded because the destination
958 address was not a local address [ipv6IfStatsInAddr‐
959 Errors].
960 iukwnp6/s
962 The number of locally-addressed datagrams received suc‐
963 cessfully but discarded per second because of an unknown
964 or unsupported protocol [ipv6IfStatsInUnknownProtos].
965 i2big6/s
967 The number of input datagrams that could not be forwarded
968 per second because their size exceeded the link MTU of
969 outgoing interface [ipv6IfStatsInTooBigErrors].
970 idisc6/s
972 The number of input IPv6 datagrams per second for which
973 no problems were encountered to prevent their continued
974 processing, but which were discarded (e.g., for lack of
975 buffer space) [ipv6IfStatsInDiscards]. Note that this
976 counter does not include any datagrams discarded while
977 awaiting re-assembly.
978 odisc6/s
980 The number of output IPv6 datagrams per second for which
981 no problem was encountered to prevent their transmission
982 to their destination, but which were discarded (e.g., for
983 lack of buffer space) [ipv6IfStatsOutDiscards]. Note that
984 this counter would include datagrams counted in fwddgm6/s
985 if any such packets met this (discretionary) discard cri‐
986 terion.
987 inort6/s
989 The number of input datagrams discarded per second
990 because no route could be found to transmit them to their
991 destination [ipv6IfStatsInNoRoutes].
992 onort6/s
994 The number of locally generated IP datagrams discarded
995 per second because no route could be found to transmit
996 them to their destination [unknown formal SNMP name].
997 asmf6/s
999 The number of failures detected per second by the IPv6
1000 re-assembly algorithm (for whatever reason: timed out,
1001 errors, etc.) [ipv6IfStatsReasmFails]. Note that this is
1002 not necessarily a count of discarded IPv6 fragments since
1003 some algorithms can lose track of the number of fragments
1004 by combining them as they are received.
1005 fragf6/s
1007 The number of IPv6 datagrams that have been discarded per
1008 second because they needed to be fragmented at this out‐
1009 put interface but could not be [ipv6IfStatsOutFragFails].
1010 itrpck6/s
1012 The number of input datagrams discarded per second
1013 because datagram frame didn't carry enough data [ipv6IfS‐
1014 tatsInTruncatedPkts].
1015 With the NFS keyword, statistics about NFS client activity are
1017 reported. The following values are displayed:
1018 call/s
1020 Number of RPC requests made per second.
1021 retrans/s
1023 Number of RPC requests per second, those which needed to
1024 be retransmitted (for example because of a server time‐
1025 out).
1026 read/s
1028 Number of 'read' RPC calls made per second.
1029 write/s
1031 Number of 'write' RPC calls made per second.
1032 access/s
1034 Number of 'access' RPC calls made per second.
1035 getatt/s
1037 Number of 'getattr' RPC calls made per second.
1038 With the NFSD keyword, statistics about NFS server activity are
1040 reported. The following values are displayed:
1041 scall/s
1043 Number of RPC requests received per second.
1044 badcall/s
1046 Number of bad RPC requests received per second, those
1047 whose processing generated an error.
1048 packet/s
1050 Number of network packets received per second.
1051 udp/s
1053 Number of UDP packets received per second.
1054 tcp/s
1056 Number of TCP packets received per second.
1057 hit/s
1059 Number of reply cache hits per second.
1060 miss/s
1062 Number of reply cache misses per second.
1063 sread/s
1065 Number of 'read' RPC calls received per second.
1066 swrite/s
1068 Number of 'write' RPC calls received per second.
1069 saccess/s
1071 Number of 'access' RPC calls received per second.
1072 sgetatt/s
1074 Number of 'getattr' RPC calls received per second.
1075 With the SOCK keyword, statistics on sockets in use are reported
1077 (IPv4). The following values are displayed:
1078 totsck
1080 Total number of sockets used by the system.
1081 tcpsck
1083 Number of TCP sockets currently in use.
1084 udpsck
1086 Number of UDP sockets currently in use.
1087 rawsck
1089 Number of RAW sockets currently in use.
1090 ip-frag
1092 Number of IP fragments currently in queue.
1093 tcp-tw
1095 Number of TCP sockets in TIME_WAIT state.
1096 With the SOCK6 keyword, statistics on sockets in use are
1098 reported (IPv6). Note that IPv6 statistics depend on sadc's
1099 option "-S IPV6" to be collected. The following values are dis‐
1100 played:
1101 tcp6sck
1103 Number of TCPv6 sockets currently in use.
1104 udp6sck
1106 Number of UDPv6 sockets currently in use.
1107 raw6sck
1109 Number of RAWv6 sockets currently in use.
1110 ip6-frag
1112 Number of IPv6 fragments currently in use.
1113 With the SOFT keyword, statistics about software-based network
1115 processing are reported. The following values are displayed:
1116 total/s
1118 The total number of network frames processed per second.
1119 dropd/s
1121 The total number of network frames dropped per second
1122 because there was no room on the processing queue.
1123 squeezd/s
1125 The number of times the softirq handler function termi‐
1126 nated per second because its budget was consumed or the
1127 time limit was reached, but more work could have been
1128 done.
1129 rx_rps/s
1131 The number of times the CPU has been woken up per second
1132 to process packets via an inter-processor interrupt.
1133 flw_lim/s
1135 The number of times the flow limit has been reached per
1136 second. Flow limiting is an optional RPS feature that
1137 can be used to limit the number of packets queued to the
1138 backlog for each flow to a certain amount. This can help
1139 ensure that smaller flows are processed even though much
1140 larger flows are pushing packets in.
1141 With the TCP keyword, statistics about TCPv4 network traffic are
1143 reported. Note that TCPv4 statistics depend on sadc's option
1144 "-S SNMP" to be collected. The following values are displayed
1145 (formal SNMP names between square brackets):
1146 active/s
1148 The number of times TCP connections have made a direct
1149 transition to the SYN-SENT state from the CLOSED state
1150 per second [tcpActiveOpens].
1151 passive/s
1153 The number of times TCP connections have made a direct
1154 transition to the SYN-RCVD state from the LISTEN state
1155 per second [tcpPassiveOpens].
1156 iseg/s
1158 The total number of segments received per second, includ‐
1159 ing those received in error [tcpInSegs]. This count
1160 includes segments received on currently established con‐
1161 nections.
1162 oseg/s
1164 The total number of segments sent per second, including
1165 those on current connections but excluding those contain‐
1166 ing only retransmitted octets [tcpOutSegs].
1167 With the ETCP keyword, statistics about TCPv4 network errors are
1169 reported. Note that TCPv4 statistics depend on sadc's option
1170 "-S SNMP" to be collected. The following values are displayed
1171 (formal SNMP names between square brackets):
1172 atmptf/s
1174 The number of times per second TCP connections have made
1175 a direct transition to the CLOSED state from either the
1176 SYN-SENT state or the SYN-RCVD state, plus the number of
1177 times per second TCP connections have made a direct tran‐
1178 sition to the LISTEN state from the SYN-RCVD state
1179 [tcpAttemptFails].
1180 estres/s
1182 The number of times per second TCP connections have made
1183 a direct transition to the CLOSED state from either the
1184 ESTABLISHED state or the CLOSE-WAIT state [tcpEstabRe‐
1185 sets].
1186 retrans/s
1188 The total number of segments retransmitted per second -
1189 that is, the number of TCP segments transmitted contain‐
1190 ing one or more previously transmitted octets [tcpRe‐
1191 transSegs].
1192 isegerr/s
1194 The total number of segments received in error (e.g., bad
1195 TCP checksums) per second [tcpInErrs].
1196 orsts/s
1198 The number of TCP segments sent per second containing the
1199 RST flag [tcpOutRsts].
1200 With the UDP keyword, statistics about UDPv4 network traffic are
1202 reported. Note that UDPv4 statistics depend on sadc's option
1203 "-S SNMP" to be collected. The following values are displayed
1204 (formal SNMP names between square brackets):
1205 idgm/s
1207 The total number of UDP datagrams delivered per second to
1208 UDP users [udpInDatagrams].
1209 odgm/s
1211 The total number of UDP datagrams sent per second from
1212 this entity [udpOutDatagrams].
1213 noport/s
1215 The total number of received UDP datagrams per second for
1216 which there was no application at the destination port
1217 [udpNoPorts].
1218 idgmerr/s
1220 The number of received UDP datagrams per second that
1221 could not be delivered for reasons other than the lack of
1222 an application at the destination port [udpInErrors].
1223 With the UDP6 keyword, statistics about UDPv6 network traffic
1225 are reported. Note that UDPv6 statistics depend on sadc's
1226 option "-S IPV6" to be collected. The following values are dis‐
1227 played (formal SNMP names between square brackets):
1228 idgm6/s
1230 The total number of UDP datagrams delivered per second to
1231 UDP users [udpInDatagrams].
1232 odgm6/s
1234 The total number of UDP datagrams sent per second from
1235 this entity [udpOutDatagrams].
1236 noport6/s
1238 The total number of received UDP datagrams per second for
1239 which there was no application at the destination port
1240 [udpNoPorts].
1241 idgmer6/s
1243 The number of received UDP datagrams per second that
1244 could not be delivered for reasons other than the lack of
1245 an application at the destination port [udpInErrors].
1246 The ALL keyword is equivalent to specifying all the keywords
1248 above and therefore all the network activities are reported.
1249 -o [ filename ]
1251 Save the readings in the file in binary form. Each reading is in
1252 a separate record. The default value of the filename parameter
1253 is the current standard system activity daily data file. If
1254 filename is a directory instead of a plain file then it is con‐
1255 sidered as the directory where the standard system activity
1256 daily data files are located. The -o option is exclusive of the
1257 -f option. All the data available from the kernel are saved in
1258 the file (in fact, sar calls its data collector sadc with the
1259 option "-S ALL". See sadc(8) manual page).
1260 -P { cpu_list | ALL }
1262 Report per-processor statistics for the specified processor or
1263 processors. cpu_list is a list of comma-separated values or
1264 range of values (e.g., 0,2,4-7,12-). Note that processor 0 is
1265 the first processor, and processor all is the global average
1266 among all processors. Specifying the ALL keyword reports sta‐
1267 tistics for each individual processor, and globally for all pro‐
1268 cessors. Offline processors are not displayed.
1269 -p Pretty-print device names. Use this option in conjunction with
1271 option -d. By default names are printed as devM-n where M and n
1272 are the major and minor numbers for the device. Use of this
1273 option displays the names of the devices as they (should) appear
1274 in /dev. Name mappings are controlled by /etc/sysconfig/sys‐
1275 stat.ioconf.
1276 -q Report queue length and load averages. The following values are
1278 displayed:
1279 runq-sz
1281 Run queue length (number of tasks waiting for run time).
1282 plist-sz
1284 Number of tasks in the task list.
1285 ldavg-1
1287 System load average for the last minute. The load aver‐
1288 age is calculated as the average number of runnable or
1289 running tasks (R state), and the number of tasks in unin‐
1290 terruptible sleep (D state) over the specified interval.
1291 ldavg-5
1293 System load average for the past 5 minutes.
1294 ldavg-15
1296 System load average for the past 15 minutes.
1297 blocked
1299 Number of tasks currently blocked, waiting for I/O to
1300 complete.
1301 -r [ ALL ]
1303 Report memory utilization statistics. The ALL keyword indicates
1304 that all the memory fields should be displayed. The following
1305 values may be displayed:
1306 kbmemfree
1308 Amount of free memory available in kilobytes.
1309 kbavail
1311 Estimate of how much memory in kilobytes is available for
1312 starting new applications, without swapping. The esti‐
1313 mate takes into account that the system needs some page
1314 cache to function well, and that not all reclaimable slab
1315 will be reclaimable, due to items being in use. The
1316 impact of those factors will vary from system to system.
1317 kbmemused
1319 Amount of used memory in kilobytes. This does not take
1320 into account memory used by the kernel itself.
1321 %memused
1323 Percentage of used memory.
1324 kbbuffers
1326 Amount of memory used as buffers by the kernel in kilo‐
1327 bytes.
1328 kbcached
1330 Amount of memory used to cache data by the kernel in
1331 kilobytes.
1332 kbcommit
1334 Amount of memory in kilobytes needed for current work‐
1335 load. This is an estimate of how much RAM/swap is needed
1336 to guarantee that there never is out of memory.
1337 %commit
1339 Percentage of memory needed for current workload in rela‐
1340 tion to the total amount of memory (RAM+swap). This num‐
1341 ber may be greater than 100% because the kernel usually
1342 overcommits memory.
1343 kbactive
1345 Amount of active memory in kilobytes (memory that has
1346 been used more recently and usually not reclaimed unless
1347 absolutely necessary).
1348 kbinact
1350 Amount of inactive memory in kilobytes (memory which has
1351 been less recently used. It is more eligible to be
1352 reclaimed for other purposes).
1353 kbdirty
1355 Amount of memory in kilobytes waiting to get written back
1356 to the disk.
1357 kbanonpg
1359 Amount of non-file backed pages in kilobytes mapped into
1360 userspace page tables.
1361 kbslab
1363 Amount of memory in kilobytes used by the kernel to cache
1364 data structures for its own use.
1365 kbkstack
1367 Amount of memory in kilobytes used for kernel stack
1368 space.
1369 kbpgtbl
1371 Amount of memory in kilobytes dedicated to the lowest
1372 level of page tables.
1373 kbvmused
1375 Amount of memory in kilobytes of used virtual address
1376 space.
1377 -S Report swap space utilization statistics. The following values
1379 are displayed:
1380 kbswpfree
1382 Amount of free swap space in kilobytes.
1383 kbswpused
1385 Amount of used swap space in kilobytes.
1386 %swpused
1388 Percentage of used swap space.
1389 kbswpcad
1391 Amount of cached swap memory in kilobytes. This is mem‐
1392 ory that once was swapped out, is swapped back in but
1393 still also is in the swap area (if memory is needed it
1394 doesn't need to be swapped out again because it is
1395 already in the swap area. This saves I/O).
1396 %swpcad
1398 Percentage of cached swap memory in relation to the
1399 amount of used swap space.
1400 -s [ hh:mm[:ss] ]
1402 Set the starting time of the data, causing the sar command to
1403 extract records time-tagged at, or following, the time speci‐
1404 fied. The default starting time is 08:00:00. Hours must be
1405 given in 24-hour format. This option can be used only when data
1406 are read from a file (option -f).
1407 --sadc Indicate which data collector is called by sar. If the data
1409 collector is sought in PATH then enter "which sadc" to know
1410 where it is located.
1411 -t When reading data from a daily data file, indicate that sar
1413 should display the timestamps in the original local time of the
1414 data file creator. Without this option, the sar command displays
1415 the timestamps in the user's locale time.
1416 -u [ ALL ]
1418 Report CPU utilization. The ALL keyword indicates that all the
1419 CPU fields should be displayed. The report may show the follow‐
1420 ing fields:
1421 %user
1423 Percentage of CPU utilization that occurred while execut‐
1424 ing at the user level (application). Note that this field
1425 includes time spent running virtual processors.
1426 %usr
1428 Percentage of CPU utilization that occurred while execut‐
1429 ing at the user level (application). Note that this field
1430 does NOT include time spent running virtual processors.
1431 %nice
1433 Percentage of CPU utilization that occurred while execut‐
1434 ing at the user level with nice priority.
1435 %system
1437 Percentage of CPU utilization that occurred while execut‐
1438 ing at the system level (kernel). Note that this field
1439 includes time spent servicing hardware and software
1440 interrupts.
1441 %sys
1443 Percentage of CPU utilization that occurred while execut‐
1444 ing at the system level (kernel). Note that this field
1445 does NOT include time spent servicing hardware or soft‐
1446 ware interrupts.
1447 %iowait
1449 Percentage of time that the CPU or CPUs were idle during
1450 which the system had an outstanding disk I/O request.
1451 %steal
1453 Percentage of time spent in involuntary wait by the vir‐
1454 tual CPU or CPUs while the hypervisor was servicing
1455 another virtual processor.
1456 %irq
1458 Percentage of time spent by the CPU or CPUs to service
1459 hardware interrupts.
1460 %soft
1462 Percentage of time spent by the CPU or CPUs to service
1463 software interrupts.
1464 %guest
1466 Percentage of time spent by the CPU or CPUs to run a vir‐
1467 tual processor.
1468 %gnice
1470 Percentage of time spent by the CPU or CPUs to run a
1471 niced guest.
1472 %idle
1474 Percentage of time that the CPU or CPUs were idle and the
1475 system did not have an outstanding disk I/O request.
1476 -V Print version number then exit.
1478 -v Report status of inode, file and other kernel tables. The fol‐
1480 lowing values are displayed:
1481 dentunusd
1483 Number of unused cache entries in the directory cache.
1484 file-nr
1486 Number of file handles used by the system.
1487 inode-nr
1489 Number of inode handlers used by the system.
1490 pty-nr
1492 Number of pseudo-terminals used by the system.
1493 -W Report swapping statistics. The following values are displayed:
1495 pswpin/s
1497 Total number of swap pages the system brought in per sec‐
1498 ond.
1499 pswpout/s
1501 Total number of swap pages the system brought out per
1502 second.
1503 -w Report task creation and system switching activity.
1505 proc/s
1507 Total number of tasks created per second.
1508 cswch/s
1510 Total number of context switches per second.
1511 -y Report TTY devices activity. The following values are displayed:
1513 rcvin/s
1515 Number of receive interrupts per second for current
1516 serial line. Serial line number is given in the TTY col‐
1517 umn.
1518 xmtin/s
1520 Number of transmit interrupts per second for current
1521 serial line.
1522 framerr/s
1524 Number of frame errors per second for current serial
1525 line.
1526 prtyerr/s
1528 Number of parity errors per second for current serial
1529 line.
1530 brk/s
1532 Number of breaks per second for current serial line.
1533 ovrun/s
1535 Number of overrun errors per second for current serial
1536 line.
1537 -z Tell sar to omit output for any devices for which there was no
1539 activity during the sample period.
1540
1543 The sar command takes into account the following environment variables:
1544 S_COLORS
1547 When this variable is set, display statistics in color on the
1548 terminal. Possible values for this variable are never, always
1549 or auto (the latter is the default).
1550 Please note that the color (being red, yellow, or some other
1552 color) used to display a value is not indicative of any kind of
1553 issue simply because of the color. It only indicates different
1554 ranges of values.
1555 S_COLORS_SGR
1558 Specify the colors and other attributes used to display statis‐
1559 tics on the terminal. Its value is a colon-separated list of
1560 capabilities that defaults to
1561 C=33;22:H=31;1:I=32;22:M=35;1:N=34;1:R=31;22:Z=34;22. Supported
1562 capabilities are:
1563 C= SGR (Select Graphic Rendition) substring for comments
1566 inserted in the binary daily data files.
1567 H= SGR substring for percentage values greater than or equal
1570 to 75%.
1571 I= SGR substring for item names or values (eg. network
1574 interfaces, CPU number...)
1575 M= SGR substring for percentage values in the range from 50%
1578 to 75%.
1579 N= SGR substring for non-zero statistics values.
1582 R= SGR substring for restart messages.
1585 Z= SGR substring for zero values.
1588 S_TIME_DEF_TIME
1591 If this variable exists and its value is UTC then sar will save
1592 its data in UTC time (data will still be displayed in local
1593 time). sar will also use UTC time instead of local time to
1594 determine the current daily data file located in the /var/log/sa
1595 directory. This variable may be useful for servers with users
1596 located across several timezones.
1597 S_TIME_FORMAT
1600 If this variable exists and its value is ISO then the current
1601 locale will be ignored when printing the date in the report
1602 header. The sar command will use the ISO 8601 format (YYYY-MM-
1603 DD) instead. The timestamp will also be compliant with ISO 8601
1604 format.
1605
1607 sar -u 2 5
1608 Report CPU utilization for each 2 seconds. 5 lines are dis‐
1609 played.
1610 sar -I 14 -o int14.file 2 10
1612 Report statistics on IRQ 14 for each 2 seconds. 10 lines are
1613 displayed. Data are stored in a file called int14.file.
1614 sar -r -n DEV -f /var/log/sa/sa16
1616 Display memory and network statistics saved in daily data file
1617 'sa16'.
1618 sar -A
1620 Display all the statistics saved in current daily data file.
1621
1623 /proc filesystem must be mounted for the sar command to work.
1624 All the statistics are not necessarily available, depending on the ker‐
1626 nel version used. sar assumes that you are using at least a 2.6 ker‐
1627 nel.
1628
1630 /var/log/sa/saDD
1631 /var/log/sa/saYYYYMMDD
1632 The standard system activity daily data files and their default
1633 location. YYYY stands for the current year, MM for the current
1634 month and DD for the current day.
1635 /proc and /sys contain various files with system statistics.
1637
1639 Sebastien Godard (sysstat <at> orange.fr)
1640
1642 sadc(8), sa1(8), sa2(8), sadf(1), sysstat(5), pidstat(1), mpstat(1),
1643 iostat(1), vmstat(8)
1644 http://pagesperso-orange.fr/sebastien.godard/
1646Linux FEBRUARY 2018 SAR(1)