1 .TH SAR 1 "MAY 2018" Linux "Linux User's Manual" -*- nroff -*-
3 sar \- Collect, report, or save system activity information.
5 .B sar [ -A ] [ -B ] [ -b ] [ -C ] [ -D ] [ -d ] [ -F [ MOUNT ] ] [ -H ] [ -h ] [ -p ] [ -q ]
6 .B [ -r [ ALL ] ] [ -S ] [ -t ] [ -u [ ALL ] ] [ -V ] [ -v ] [ -W ] [ -w ] [ -y ] [ -z ]
7 .B [ --dec={ 0 | 1 | 2 } ] [ --dev=
11 .B ] [ --help ] [ --human ] [ --iface=
16 .B | SUM | ALL } ] [ -P {
25 .B [ -j { ID | LABEL | PATH | UUID | ... } ]
45 command writes to standard output the contents of selected
46 cumulative activity counters in the operating system. The accounting
47 system, based on the values in the
51 parameters, writes information the specified number of times spaced
52 at the specified intervals in seconds.
55 parameter is set to zero, the
57 command displays the average statistics for the time
58 since the system was started. If the
60 parameter is specified without the
62 parameter, then reports are generated continuously.
63 The collected data can also
64 be saved in the file specified by the -o
66 flag, in addition to being displayed onto the screen. If
70 uses the standard system activity daily data file (see below).
71 By default all the data available from the kernel are saved in the
76 command extracts and writes to standard output records previously
77 saved in a file. This file can be either the one specified by the
78 -f flag or, by default, the standard system activity daily data file.
79 It is also possible to enter -1, -2 etc. as an argument to
82 of that days ago. For example, -1 will point at the standard system
83 activity file of yesterday.
85 Standard system activity daily data files are named
89 where YYYY stands for the current year, MM for the current month and
90 DD for the current day. They are the default files used by
92 only when no filename has been explicitly specified.
93 When used to write data to files (with its option -o),
97 if option -D has also been specified, else it will use
99 When used to display the records previously saved in a file,
101 will look for the most recent of
107 Standard system activity daily data files are located in the
109 directory by default. Yet it is possible to specify an alternate
110 location for them: If a directory (instead of a plain file) is used
111 with options -f or -o
112 then it will be considered as the directory containing the data files.
114 Without the -P flag, the
116 command reports system-wide (global among all processors) statistics,
117 which are calculated as averages for values expressed as percentages,
118 and as sums otherwise. If the -P
121 command reports activity which relates to the specified processor or
122 processors. If -P ALL
125 command reports statistics for each individual processor and global
126 statistics among all processors. Offline processors are not displayed.
128 You can select information about specific system activities using
129 flags. Not specifying any flags selects only CPU activity.
131 flag selects all possible activities.
133 The default version of the
135 command (CPU utilization report) might be one of the first facilities
136 the user runs to begin system activity investigation, because it
137 monitors major system resources. If CPU utilization is near 100 percent
138 (user + nice + system), the workload sampled is CPU-bound.
140 If multiple samples and multiple reports are desired, it is convenient
141 to specify an output file for the
146 command as a background process. The syntax for this is:
148 .B sar -o datafile interval count >/dev/null 2>&1 &
150 All data are captured in binary form and saved to a file (datafile).
151 The data can then be selectively displayed with the
162 second intervals. If the
164 parameter is not set, all the records saved in the
165 file will be selected.
166 Collection of data in this manner is useful to characterize
167 system usage over a period of time and determine peak usage hours.
171 command only reports on local activities.
175 This is equivalent to specifying
176 .BR "-bBdFHqSuvwWy -I SUM -I ALL -m ALL -n ALL -r ALL -u ALL -P ALL".
178 Report paging statistics.
179 The following values are displayed:
184 Total number of kilobytes the system paged in from disk per second.
189 Total number of kilobytes the system paged out to disk per second.
194 Number of page faults (major + minor) made by the system per second.
195 This is not a count of page faults that generate I/O, because some page
196 faults can be resolved without I/O.
201 Number of major faults the system has made per second, those which
202 have required loading a memory page from disk.
207 Number of pages placed on the free list by the system per second.
212 Number of pages scanned by the kswapd daemon per second.
217 Number of pages scanned directly per second.
222 Number of pages the system has reclaimed from cache (pagecache and
223 swapcache) per second to satisfy its memory demands.
228 Calculated as pgsteal / pgscan, this is a metric of the efficiency of
229 page reclaim. If it is near 100% then almost every page coming off the
230 tail of the inactive list is being reaped. If it gets too low (e.g. less
231 than 30%) then the virtual memory is having some difficulty.
232 This field is displayed as zero if no pages have been scanned during the
237 Report I/O and transfer rate statistics.
238 The following values are displayed:
243 Total number of transfers per second that were issued to physical devices.
244 A transfer is an I/O request to a physical device. Multiple logical
245 requests can be combined into a single I/O request to the device.
246 A transfer is of indeterminate size.
251 Total number of read requests per second issued to physical devices.
256 Total number of write requests per second issued to physical devices.
261 Total amount of data read from the devices in blocks per second.
262 Blocks are equivalent to sectors
263 and therefore have a size of 512 bytes.
268 Total amount of data written to devices in blocks per second.
272 When reading data from a file, tell
274 to display comments that have been inserted by
281 as the standard system activity daily data file name. This option
282 works only when used in conjunction with option -o
283 to save data to file.
285 Report activity for each block device.
286 When data are displayed, the device specification
288 is generally used (DEV column).
289 M is the major number of the device and n
291 Device names may also be pretty-printed if option -p
292 is used or persistent device names can be printed if option -j is used
293 (see below). Statistics for all devices are displayed unless
294 a restricted list is specified using option
296 (see corresponding option entry).
297 Note that disk activity depends on
303 to be collected. The following values are displayed:
308 Total number of transfers per second that were issued to physical devices.
309 A transfer is an I/O request to a physical device. Multiple logical
310 requests can be combined into a single I/O request to the device.
311 A transfer is of indeterminate size.
316 Number of kilobytes read from the device per second.
321 Number of kilobytes written to the device per second.
326 The average size (in kilobytes) of the I/O requests that were issued to the device.
328 Note: In previous versions, this field was known as avgrq-sz and was expressed in sectors.
333 The average queue length of the requests that were issued to the device.
335 Note: In previous versions, this field was known as avgqu-sz.
340 The average time (in milliseconds) for I/O requests issued to the device
341 to be served. This includes the time spent by the requests in queue and
342 the time spent servicing them.
347 The average service time (in milliseconds) for I/O requests that were issued
348 to the device. Warning! Do not trust this field any more. This field will be
349 removed in a future sysstat version.
354 Percentage of elapsed time during which I/O requests were issued to the device
355 (bandwidth utilization for the device). Device saturation occurs when this
356 value is close to 100% for devices serving requests serially. But for
357 devices serving requests in parallel, such as RAID arrays and modern SSDs,
358 this number does not reflect their performance limits.
361 .IP "--dec={ 0 | 1 | 2 }"
362 Specify the number of decimal places to use (0 to 2, default value is 2).
364 Specify the block devices for which statistics are to be displayed by
367 is a list of comma-separated device names.
368 .IP "-e [ hh:mm[:ss] ]"
369 Set the ending time of the report. The default ending time is
370 18:00:00. Hours must be given in 24-hour format.
371 This option can be used when data are read from
372 or written to a file (options -f or -o).
374 Display statistics for currently mounted filesystems. Pseudo-filesystems are
375 ignored. At the end of the report,
377 will display a summary of all those filesystems.
380 parameter keyword indicates that mountpoint will be reported instead of
381 filesystem device. Statistics for all filesystems are displayed unless
382 a restricted list is specified using option
384 (see corresponding option entry).
385 Note that filesystems statistics depend on
391 The following values are displayed:
396 Total amount a free space in megabytes (including space available only to privileged user).
401 Total amount of space used in megabytes.
406 Percentage of filesystem space used, as seen by a privileged user.
411 Percentage of filesystem space used, as seen by an unprivileged user.
416 Total number of free file nodes in filesystem.
421 Total number of file nodes used in filesystem.
426 Percentage of file nodes used in filesystem.
429 .IP "-f [ filename ]"
434 flag). The default value of the
436 parameter is the current standard system activity daily data file.
439 is a directory instead of a plain file then it is considered as the
440 directory where the standard system activity daily data files are
441 located. The -f option is exclusive of the -o option.
443 Specify the filesystems for which statistics are to be displayed by
446 is a list of comma-separated filesystem names or mountpoints.
448 Report hugepages utilization statistics.
449 The following values are displayed:
454 Amount of hugepages memory in kilobytes that is not yet allocated.
459 Amount of hugepages memory in kilobytes that has been allocated.
464 Percentage of total hugepages memory that has been allocated.
468 Make the output of sar easier to read by a human. Options
472 (pretty-print) are enabled implicitly with this option.
474 Display a short help message then exit.
476 Print sizes in human readable format (e.g. 1.0k, 1.2M, etc.)
477 The units displayed with this option supersede any other default units (e.g.
478 kilobytes, sectors...) associated with the metrics.
479 .IP "-I { int_list | SUM | ALL }"
480 Report statistics for interrupts.
482 is a list of comma-separated values or range of values (e.g.,
486 keyword indicates that the total number of interrupts received per second
487 is to be displayed. The
489 keyword indicates that statistics from all interrupts, including potential
490 APIC interrupt sources, are to be reported.
491 Note that interrupt statistics depend on
496 Select data records at seconds as close as possible to the number specified
500 .IP --iface=iface_list
501 Specify the network interfaces for which statistics are to be displayed by
504 is a list of comma-separated interface names.
505 .IP "-j { ID | LABEL | PATH | UUID | ... }"
506 Display persistent device names. Use this option in conjunction with option -d.
510 etc. specify the type of the persistent name. These options are not limited,
511 only prerequisite is that directory with required persistent names is present in
513 If persistent name is not found for the device, the device name
514 is pretty-printed (see option -p below).
515 .IP "-m { keyword [,...] | ALL }"
516 Report power management statistics.
517 Note that these statistics depend on
519 option "-S POWER" to be collected.
521 Possible keywords are
532 keyword, statistics about CPU are reported.
533 The following value is displayed:
538 Instantaneous CPU clock frequency in MHz.
543 keyword, statistics about fans speed are reported.
544 The following values are displayed:
548 Fan speed expressed in revolutions per minute.
553 This field is calculated as the difference between current fan speed (rpm)
554 and its low limit (fan_min).
564 keyword, statistics about CPU clock frequency are reported.
565 The following value is displayed:
569 Weighted average CPU clock frequency in MHz.
570 Note that the cpufreq-stats driver must be compiled in the
571 kernel for this option to work.
576 keyword, statistics about voltage inputs are reported.
577 The following values are displayed:
581 Voltage input expressed in Volts.
586 Relative input value. A value of 100% means that
587 voltage input has reached its high limit (in_max) whereas
588 a value of 0% means that it has reached its low limit (in_min).
598 keyword, statistics about devices temperature are reported.
599 The following values are displayed:
603 Device temperature expressed in degrees Celsius.
608 Relative device temperature. A value of 100% means that
609 temperature has reached its high limit (temp_max).
621 command takes a snapshot of all the USB devices currently plugged into
622 the system. At the end of the report,
624 will display a summary of all those USB devices.
625 The following values are displayed:
629 Root hub number of the USB device.
634 Vendor ID number (assigned by USB organization).
639 Product ID number (assigned by Manufacturer).
644 Maximum power consumption of the device (expressed in mA).
659 keyword is equivalent to specifying all the keywords above and therefore all the power
660 management statistics are reported.
663 .IP "-n { keyword [,...] | ALL }"
664 Report network statistics.
666 Possible keywords are
691 keyword, statistics from the network devices are reported.
692 Statistics for all network interfaces are displayed unless
693 a restricted list is specified using option
695 (see corresponding option entry).
696 The following values are displayed:
701 Name of the network interface for which statistics are reported.
706 Total number of packets received per second.
711 Total number of packets transmitted per second.
716 Total number of kilobytes received per second.
721 Total number of kilobytes transmitted per second.
726 Number of compressed packets received per second (for cslip etc.).
731 Number of compressed packets transmitted per second.
736 Number of multicast packets received per second.
741 Utilization percentage of the network interface. For half-duplex interfaces,
742 utilization is calculated using the sum of rxkB/s and txkB/s as a percentage
743 of the interface speed. For full-duplex, this is the greater of rxkB/S or txkB/s.
748 keyword, statistics on failures (errors) from the network devices are reported.
749 Statistics for all network interfaces are displayed unless
750 a restricted list is specified using option
752 (see corresponding option entry).
753 The following values are displayed:
757 Name of the network interface for which statistics are reported.
762 Total number of bad packets received per second.
767 Total number of errors that happened per second while transmitting packets.
772 Number of collisions that happened per second while transmitting packets.
777 Number of received packets dropped per second because of a lack of space in linux buffers.
782 Number of transmitted packets dropped per second because of a lack of space in linux buffers.
787 Number of carrier-errors that happened per second while transmitting packets.
792 Number of frame alignment errors that happened per second on received packets.
797 Number of FIFO overrun errors that happened per second on received packets.
802 Number of FIFO overrun errors that happened per second on transmitted packets.
807 keyword, statistics about fibre channel traffic are reported.
808 Note that fibre channel statistics depend on
810 option "-S DISK" to be collected.
811 The following values are displayed:
815 Name of the fibre channel host bus adapter (HBA) interface for which statistics are reported.
820 The total number of frames received per second.
825 The total number of frames transmitted per second.
830 The total number of transmission words received per second.
835 The total number of transmission words transmitted per second.
840 keyword, statistics about ICMPv4 network traffic are reported.
841 Note that ICMPv4 statistics depend on
845 The following values are displayed (formal SNMP names between
850 The total number of ICMP messages which the entity
851 received per second [icmpInMsgs].
852 Note that this counter includes all those counted by ierr/s.
857 The total number of ICMP messages which this entity
858 attempted to send per second [icmpOutMsgs].
859 Note that this counter includes all those counted by oerr/s.
864 The number of ICMP Echo (request) messages received per second [icmpInEchos].
869 The number of ICMP Echo Reply messages received per second [icmpInEchoReps].
874 The number of ICMP Echo (request) messages sent per second [icmpOutEchos].
879 The number of ICMP Echo Reply messages sent per second [icmpOutEchoReps].
884 The number of ICMP Timestamp (request) messages received per second [icmpInTimestamps].
889 The number of ICMP Timestamp Reply messages received per second [icmpInTimestampReps].
894 The number of ICMP Timestamp (request) messages sent per second [icmpOutTimestamps].
899 The number of ICMP Timestamp Reply messages sent per second [icmpOutTimestampReps].
904 The number of ICMP Address Mask Request messages received per second [icmpInAddrMasks].
909 The number of ICMP Address Mask Reply messages received per second [icmpInAddrMaskReps].
914 The number of ICMP Address Mask Request messages sent per second [icmpOutAddrMasks].
919 The number of ICMP Address Mask Reply messages sent per second [icmpOutAddrMaskReps].
924 keyword, statistics about ICMPv4 error messages are reported.
925 Note that ICMPv4 statistics depend on
927 option "-S SNMP" to be collected.
928 The following values are displayed (formal SNMP names between
933 The number of ICMP messages per second which the entity received but
934 determined as having ICMP-specific errors (bad ICMP
935 checksums, bad length, etc.) [icmpInErrors].
940 The number of ICMP messages per second which this entity did not send
941 due to problems discovered within ICMP such as a lack of buffers [icmpOutErrors].
946 The number of ICMP Destination Unreachable messages
947 received per second [icmpInDestUnreachs].
952 The number of ICMP Destination Unreachable messages sent per second [icmpOutDestUnreachs].
957 The number of ICMP Time Exceeded messages received per second [icmpInTimeExcds].
962 The number of ICMP Time Exceeded messages sent per second [icmpOutTimeExcds].
967 The number of ICMP Parameter Problem messages received per second [icmpInParmProbs].
972 The number of ICMP Parameter Problem messages sent per second [icmpOutParmProbs].
977 The number of ICMP Source Quench messages received per second [icmpInSrcQuenchs].
982 The number of ICMP Source Quench messages sent per second [icmpOutSrcQuenchs].
987 The number of ICMP Redirect messages received per second [icmpInRedirects].
992 The number of ICMP Redirect messages sent per second [icmpOutRedirects].
997 keyword, statistics about ICMPv6 network traffic are reported.
998 Note that ICMPv6 statistics depend on
1000 option "-S IPV6" to be collected.
1001 The following values are displayed (formal SNMP names between
1006 The total number of ICMP messages received
1007 by the interface per second which includes all those
1008 counted by ierr6/s [ipv6IfIcmpInMsgs].
1013 The total number of ICMP messages which this
1014 interface attempted to send per second [ipv6IfIcmpOutMsgs].
1019 The number of ICMP Echo (request) messages
1020 received by the interface per second [ipv6IfIcmpInEchos].
1025 The number of ICMP Echo Reply messages received
1026 by the interface per second [ipv6IfIcmpInEchoReplies].
1031 The number of ICMP Echo Reply messages sent
1032 by the interface per second [ipv6IfIcmpOutEchoReplies].
1037 The number of ICMPv6 Group Membership Query
1038 messages received by the interface per second
1039 [ipv6IfIcmpInGroupMembQueries].
1044 The number of ICMPv6 Group Membership Response messages
1045 received by the interface per second
1046 [ipv6IfIcmpInGroupMembResponses].
1051 The number of ICMPv6 Group Membership Response
1052 messages sent per second
1053 [ipv6IfIcmpOutGroupMembResponses].
1058 The number of ICMPv6 Group Membership Reduction messages
1059 received by the interface per second
1060 [ipv6IfIcmpInGroupMembReductions].
1065 The number of ICMPv6 Group Membership Reduction
1066 messages sent per second
1067 [ipv6IfIcmpOutGroupMembReductions].
1072 The number of ICMP Router Solicit messages
1073 received by the interface per second
1074 [ipv6IfIcmpInRouterSolicits].
1079 The number of ICMP Router Solicitation messages
1080 sent by the interface per second
1081 [ipv6IfIcmpOutRouterSolicits].
1086 The number of ICMP Router Advertisement messages
1087 received by the interface per second
1088 [ipv6IfIcmpInRouterAdvertisements].
1093 The number of ICMP Neighbor Solicit messages
1094 received by the interface per second
1095 [ipv6IfIcmpInNeighborSolicits].
1100 The number of ICMP Neighbor Solicitation
1101 messages sent by the interface per second
1102 [ipv6IfIcmpOutNeighborSolicits].
1107 The number of ICMP Neighbor Advertisement
1108 messages received by the interface per second
1109 [ipv6IfIcmpInNeighborAdvertisements].
1114 The number of ICMP Neighbor Advertisement
1115 messages sent by the interface per second
1116 [ipv6IfIcmpOutNeighborAdvertisements].
1121 keyword, statistics about ICMPv6 error messages are reported.
1122 Note that ICMPv6 statistics depend on
1124 option "-S IPV6" to be collected.
1125 The following values are displayed (formal SNMP names between
1130 The number of ICMP messages per second which the interface
1131 received but determined as having ICMP-specific
1132 errors (bad ICMP checksums, bad length, etc.)
1133 [ipv6IfIcmpInErrors]
1138 The number of ICMP Destination Unreachable
1139 messages received by the interface per second
1140 [ipv6IfIcmpInDestUnreachs].
1145 The number of ICMP Destination Unreachable
1146 messages sent by the interface per second
1147 [ipv6IfIcmpOutDestUnreachs].
1152 The number of ICMP Time Exceeded messages
1153 received by the interface per second
1154 [ipv6IfIcmpInTimeExcds].
1159 The number of ICMP Time Exceeded messages sent
1160 by the interface per second
1161 [ipv6IfIcmpOutTimeExcds].
1166 The number of ICMP Parameter Problem messages
1167 received by the interface per second
1168 [ipv6IfIcmpInParmProblems].
1173 The number of ICMP Parameter Problem messages
1174 sent by the interface per second
1175 [ipv6IfIcmpOutParmProblems].
1180 The number of Redirect messages received
1181 by the interface per second
1182 [ipv6IfIcmpInRedirects].
1187 The number of Redirect messages sent by
1188 the interface by second
1189 [ipv6IfIcmpOutRedirects].
1194 The number of ICMP Packet Too Big messages
1195 received by the interface per second
1196 [ipv6IfIcmpInPktTooBigs].
1201 The number of ICMP Packet Too Big messages sent
1202 by the interface per second
1203 [ipv6IfIcmpOutPktTooBigs].
1208 keyword, statistics about IPv4 network traffic are reported.
1209 Note that IPv4 statistics depend on
1213 The following values are displayed (formal SNMP names between
1218 The total number of input datagrams received from interfaces
1219 per second, including those received in error [ipInReceives].
1224 The number of input datagrams per second, for which this entity was not
1225 their final IP destination, as a result of which an attempt
1226 was made to find a route to forward them to that final
1227 destination [ipForwDatagrams].
1232 The total number of input datagrams successfully delivered per second
1233 to IP user-protocols (including ICMP) [ipInDelivers].
1238 The total number of IP datagrams which local IP user-protocols (including ICMP)
1239 supplied per second to IP in requests for transmission [ipOutRequests].
1240 Note that this counter does not include any datagrams counted in fwddgm/s.
1245 The number of IP fragments received per second which needed to be
1246 reassembled at this entity [ipReasmReqds].
1251 The number of IP datagrams successfully re-assembled per second [ipReasmOKs].
1256 The number of IP datagrams that have been successfully
1257 fragmented at this entity per second [ipFragOKs].
1262 The number of IP datagram fragments that have been
1263 generated per second as a result of fragmentation at this entity [ipFragCreates].
1268 keyword, statistics about IPv4 network errors are reported.
1269 Note that IPv4 statistics depend on
1271 option "-S SNMP" to be collected.
1272 The following values are displayed (formal SNMP names between
1277 The number of input datagrams discarded per second due to errors in
1278 their IP headers, including bad checksums, version number
1279 mismatch, other format errors, time-to-live exceeded, errors
1280 discovered in processing their IP options, etc. [ipInHdrErrors]
1285 The number of input datagrams discarded per second because the IP
1286 address in their IP header's destination field was not a
1287 valid address to be received at this entity. This count
1288 includes invalid addresses (e.g., 0.0.0.0) and addresses of
1289 unsupported Classes (e.g., Class E). For entities which are
1290 not IP routers and therefore do not forward datagrams, this
1291 counter includes datagrams discarded because the destination
1292 address was not a local address [ipInAddrErrors].
1297 The number of locally-addressed datagrams received
1298 successfully but discarded per second because of an unknown or
1299 unsupported protocol [ipInUnknownProtos].
1304 The number of input IP datagrams per second for which no problems were
1305 encountered to prevent their continued processing, but which
1306 were discarded (e.g., for lack of buffer space) [ipInDiscards].
1307 Note that this counter does not include any datagrams discarded while
1308 awaiting re-assembly.
1313 The number of output IP datagrams per second for which no problem was
1314 encountered to prevent their transmission to their
1315 destination, but which were discarded (e.g., for lack of
1316 buffer space) [ipOutDiscards].
1317 Note that this counter would include
1318 datagrams counted in fwddgm/s if any such packets met
1319 this (discretionary) discard criterion.
1324 The number of IP datagrams discarded per second because no route could
1325 be found to transmit them to their destination [ipOutNoRoutes].
1326 Note that this counter includes any packets counted in fwddgm/s
1327 which meet this 'no-route' criterion.
1328 Note that this includes any datagrams which a host cannot route because all
1329 of its default routers are down.
1334 The number of failures detected per second by the IP re-assembly
1335 algorithm (for whatever reason: timed out, errors, etc) [ipReasmFails].
1336 Note that this is not necessarily a count of discarded IP
1337 fragments since some algorithms can lose track of the number of
1338 fragments by combining them as they are received.
1343 The number of IP datagrams that have been discarded per second because
1344 they needed to be fragmented at this entity but could not
1345 be, e.g., because their Don't Fragment flag was set [ipFragFails].
1350 keyword, statistics about IPv6 network traffic are reported.
1351 Note that IPv6 statistics depend on
1353 option "-S IPV6" to be collected.
1354 The following values are displayed (formal SNMP names between
1359 The total number of input datagrams received from
1360 interfaces per second, including those received in error
1361 [ipv6IfStatsInReceives].
1366 The number of output datagrams per second which this
1367 entity received and forwarded to their final
1368 destinations [ipv6IfStatsOutForwDatagrams].
1373 The total number of datagrams successfully
1374 delivered per second to IPv6 user-protocols (including ICMP)
1375 [ipv6IfStatsInDelivers].
1380 The total number of IPv6 datagrams which local IPv6
1381 user-protocols (including ICMP) supplied per second to IPv6 in
1382 requests for transmission [ipv6IfStatsOutRequests].
1383 Note that this counter
1384 does not include any datagrams counted in fwddgm6/s.
1389 The number of IPv6 fragments received per second which needed
1390 to be reassembled at this interface [ipv6IfStatsReasmReqds].
1395 The number of IPv6 datagrams successfully
1396 reassembled per second [ipv6IfStatsReasmOKs].
1401 The number of multicast packets received per second
1402 by the interface [ipv6IfStatsInMcastPkts].
1407 The number of multicast packets transmitted per second
1408 by the interface [ipv6IfStatsOutMcastPkts].
1413 The number of IPv6 datagrams that have been
1414 successfully fragmented at this output interface per second
1415 [ipv6IfStatsOutFragOKs].
1420 The number of output datagram fragments that have
1421 been generated per second as a result of fragmentation at
1422 this output interface [ipv6IfStatsOutFragCreates].
1427 keyword, statistics about IPv6 network errors are reported.
1428 Note that IPv6 statistics depend on
1430 option "-S IPV6" to be collected.
1431 The following values are displayed (formal SNMP names between
1436 The number of input datagrams discarded per second due to
1437 errors in their IPv6 headers, including version
1438 number mismatch, other format errors, hop count
1439 exceeded, errors discovered in processing their
1440 IPv6 options, etc. [ipv6IfStatsInHdrErrors]
1445 The number of input datagrams discarded per second because
1446 the IPv6 address in their IPv6 header's destination
1447 field was not a valid address to be received at
1448 this entity. This count includes invalid
1449 addresses (e.g., ::0) and unsupported addresses
1450 (e.g., addresses with unallocated prefixes). For
1451 entities which are not IPv6 routers and therefore
1452 do not forward datagrams, this counter includes
1453 datagrams discarded because the destination address
1454 was not a local address [ipv6IfStatsInAddrErrors].
1459 The number of locally-addressed datagrams
1460 received successfully but discarded per second because of an
1461 unknown or unsupported protocol [ipv6IfStatsInUnknownProtos].
1466 The number of input datagrams that could not be
1467 forwarded per second because their size exceeded the link MTU
1468 of outgoing interface [ipv6IfStatsInTooBigErrors].
1473 The number of input IPv6 datagrams per second for which no
1474 problems were encountered to prevent their
1475 continued processing, but which were discarded
1476 (e.g., for lack of buffer space)
1477 [ipv6IfStatsInDiscards]. Note that this
1478 counter does not include any datagrams discarded
1479 while awaiting re-assembly.
1484 The number of output IPv6 datagrams per second for which no
1485 problem was encountered to prevent their
1486 transmission to their destination, but which were
1487 discarded (e.g., for lack of buffer space)
1488 [ipv6IfStatsOutDiscards]. Note
1489 that this counter would include datagrams counted
1490 in fwddgm6/s if any such packets
1491 met this (discretionary) discard criterion.
1496 The number of input datagrams discarded per second because no
1497 route could be found to transmit them to their
1498 destination [ipv6IfStatsInNoRoutes].
1503 The number of locally generated IP datagrams discarded per second
1504 because no route could be found to transmit them to their
1505 destination [unknown formal SNMP name].
1510 The number of failures detected per second by the IPv6
1511 re-assembly algorithm (for whatever reason: timed
1512 out, errors, etc.) [ipv6IfStatsReasmFails].
1513 Note that this is not
1514 necessarily a count of discarded IPv6 fragments
1515 since some algorithms
1516 can lose track of the number of fragments
1517 by combining them as they are received.
1522 The number of IPv6 datagrams that have been
1523 discarded per second because they needed to be fragmented
1524 at this output interface but could not be
1525 [ipv6IfStatsOutFragFails].
1530 The number of input datagrams discarded per second because
1531 datagram frame didn't carry enough data
1532 [ipv6IfStatsInTruncatedPkts].
1537 keyword, statistics about NFS client activity are reported.
1538 The following values are displayed:
1542 Number of RPC requests made per second.
1547 Number of RPC requests per second, those which needed to be retransmitted (for
1548 example because of a server timeout).
1553 Number of 'read' RPC calls made per second.
1558 Number of 'write' RPC calls made per second.
1563 Number of 'access' RPC calls made per second.
1568 Number of 'getattr' RPC calls made per second.
1573 keyword, statistics about NFS server activity are reported.
1574 The following values are displayed:
1578 Number of RPC requests received per second.
1583 Number of bad RPC requests received per second, those whose
1584 processing generated an error.
1589 Number of network packets received per second.
1594 Number of UDP packets received per second.
1599 Number of TCP packets received per second.
1604 Number of reply cache hits per second.
1609 Number of reply cache misses per second.
1614 Number of 'read' RPC calls received per second.
1619 Number of 'write' RPC calls received per second.
1624 Number of 'access' RPC calls received per second.
1629 Number of 'getattr' RPC calls received per second.
1634 keyword, statistics on sockets in use are reported
1636 The following values are displayed:
1640 Total number of sockets used by the system.
1645 Number of TCP sockets currently in use.
1650 Number of UDP sockets currently in use.
1655 Number of RAW sockets currently in use.
1660 Number of IP fragments currently in queue.
1665 Number of TCP sockets in TIME_WAIT state.
1670 keyword, statistics on sockets in use are reported (IPv6).
1671 Note that IPv6 statistics depend on
1673 option "-S IPV6" to be collected.
1674 The following values are displayed:
1678 Number of TCPv6 sockets currently in use.
1683 Number of UDPv6 sockets currently in use.
1688 Number of RAWv6 sockets currently in use.
1693 Number of IPv6 fragments currently in use.
1698 keyword, statistics about software-based network processing are reported.
1699 The following values are displayed:
1703 The total number of network frames processed per second.
1708 The total number of network frames dropped per second because there
1709 was no room on the processing queue.
1714 The number of times the softirq handler function terminated per second
1715 because its budget was consumed or the time limit was reached, but more
1716 work could have been done.
1721 The number of times the CPU has been woken up per second
1722 to process packets via an inter-processor interrupt.
1727 The number of times the flow limit has been reached per second.
1728 Flow limiting is an optional RPS feature that can be used to limit the number of
1729 packets queued to the backlog for each flow to a certain amount.
1730 This can help ensure that smaller flows are processed even though
1731 much larger flows are pushing packets in.
1736 keyword, statistics about TCPv4 network traffic are reported.
1737 Note that TCPv4 statistics depend on
1739 option "-S SNMP" to be collected.
1740 The following values are displayed (formal SNMP names between
1745 The number of times TCP connections have made a direct
1746 transition to the SYN-SENT state from the CLOSED state per second [tcpActiveOpens].
1751 The number of times TCP connections have made a direct
1752 transition to the SYN-RCVD state from the LISTEN state per second [tcpPassiveOpens].
1757 The total number of segments received per second, including those
1758 received in error [tcpInSegs]. This count includes segments received on
1759 currently established connections.
1764 The total number of segments sent per second, including those on
1765 current connections but excluding those containing only
1766 retransmitted octets [tcpOutSegs].
1771 keyword, statistics about TCPv4 network errors are reported.
1772 Note that TCPv4 statistics depend on
1774 option "-S SNMP" to be collected.
1775 The following values are displayed (formal SNMP names between
1780 The number of times per second TCP connections have made a direct
1781 transition to the CLOSED state from either the SYN-SENT
1782 state or the SYN-RCVD state, plus the number of times per second TCP
1783 connections have made a direct transition to the LISTEN
1784 state from the SYN-RCVD state [tcpAttemptFails].
1789 The number of times per second TCP connections have made a direct
1790 transition to the CLOSED state from either the ESTABLISHED
1791 state or the CLOSE-WAIT state [tcpEstabResets].
1796 The total number of segments retransmitted per second - that is, the
1797 number of TCP segments transmitted containing one or more
1798 previously transmitted octets [tcpRetransSegs].
1803 The total number of segments received in error (e.g., bad
1804 TCP checksums) per second [tcpInErrs].
1809 The number of TCP segments sent per second containing the RST flag [tcpOutRsts].
1814 keyword, statistics about UDPv4 network traffic are reported.
1815 Note that UDPv4 statistics depend on
1817 option "-S SNMP" to be collected.
1818 The following values are displayed (formal SNMP names between
1823 The total number of UDP datagrams delivered per second to UDP users [udpInDatagrams].
1828 The total number of UDP datagrams sent per second from this entity [udpOutDatagrams].
1833 The total number of received UDP datagrams per second for which there
1834 was no application at the destination port [udpNoPorts].
1839 The number of received UDP datagrams per second that could not be
1840 delivered for reasons other than the lack of an application
1841 at the destination port [udpInErrors].
1846 keyword, statistics about UDPv6 network traffic are reported.
1847 Note that UDPv6 statistics depend on
1849 option "-S IPV6" to be collected.
1850 The following values are displayed (formal SNMP names between
1855 The total number of UDP datagrams delivered per second to UDP users
1861 The total number of UDP datagrams sent per second from this
1862 entity [udpOutDatagrams].
1867 The total number of received UDP datagrams per second for which there
1868 was no application at the destination port [udpNoPorts].
1873 The number of received UDP datagrams per second that could not be
1874 delivered for reasons other than the lack of an application
1875 at the destination port [udpInErrors].
1880 keyword is equivalent to specifying all the keywords above and therefore all the network
1881 activities are reported.
1884 .IP "-o [ filename ]"
1885 Save the readings in the file in binary form. Each reading
1886 is in a separate record. The default value of the
1888 parameter is the current standard system activity daily data file.
1891 is a directory instead of a plain file then it is considered as the directory
1892 where the standard system activity daily data files are located.
1893 The -o option is exclusive of the -f option.
1894 All the data available from the kernel are saved in the file (in fact,
1896 calls its data collector
1898 with the option "-S ALL".
1902 .IP "-P { cpu_list | ALL }"
1903 Report per-processor statistics for the specified processor or processors.
1905 is a list of comma-separated values or range of values (e.g.,
1907 Note that processor 0 is the first processor, and processor
1909 is the global average among all processors.
1912 keyword reports statistics for each individual processor, and globally for
1913 all processors. Offline processors are not displayed.
1915 Pretty-print device names. Use this option in conjunction with option -d.
1916 By default names are printed as
1918 where M and n are the major and minor numbers for the device.
1919 Use of this option displays the names of the devices as they (should) appear
1920 in /dev. Name mappings are controlled by
1921 .IR @SYSCONFIG_DIR@/sysstat.ioconf .
1923 Report queue length and load averages. The following values are displayed:
1928 Run queue length (number of tasks waiting for run time).
1933 Number of tasks in the task list.
1938 System load average for the last minute.
1939 The load average is calculated as the average number of runnable or
1940 running tasks (R state), and the number of tasks in uninterruptible
1941 sleep (D state) over the specified interval.
1946 System load average for the past 5 minutes.
1951 System load average for the past 15 minutes.
1956 Number of tasks currently blocked, waiting for I/O to complete.
1960 Report memory utilization statistics. The
1962 keyword indicates that all the memory fields should be displayed.
1963 The following values may be displayed:
1968 Amount of free memory available in kilobytes.
1973 Estimate of how much memory in kilobytes is available for starting new
1974 applications, without swapping.
1975 The estimate takes into account that the system needs some page cache to
1976 function well, and that not all reclaimable slab will be reclaimable,
1977 due to items being in use. The impact of those factors will vary from
1983 Amount of used memory in kilobytes (calculated as total installed memory -
1995 Percentage of used memory.
2000 Amount of memory used as buffers by the kernel in kilobytes.
2005 Amount of memory used to cache data by the kernel in kilobytes.
2010 Amount of memory in kilobytes needed for current workload. This is an estimate of how much
2011 RAM/swap is needed to guarantee that there never is out of memory.
2016 Percentage of memory needed for current workload in relation to the total amount of memory (RAM+swap).
2017 This number may be greater than 100% because the kernel usually overcommits memory.
2022 Amount of active memory in kilobytes (memory that has been used more recently
2023 and usually not reclaimed unless absolutely necessary).
2028 Amount of inactive memory in kilobytes (memory which has been less recently
2029 used. It is more eligible to be reclaimed for other purposes).
2034 Amount of memory in kilobytes waiting to get written back to the disk.
2039 Amount of non-file backed pages in kilobytes mapped into userspace page tables.
2044 Amount of memory in kilobytes used by the kernel to cache data structures for its own use.
2049 Amount of memory in kilobytes used for kernel stack space.
2054 Amount of memory in kilobytes dedicated to the lowest level of page tables.
2059 Amount of memory in kilobytes of used virtual address space.
2063 Report swap space utilization statistics.
2064 The following values are displayed:
2069 Amount of free swap space in kilobytes.
2074 Amount of used swap space in kilobytes.
2079 Percentage of used swap space.
2084 Amount of cached swap memory in kilobytes.
2085 This is memory that once was swapped out, is swapped back in
2086 but still also is in the swap area (if memory is needed it doesn't need
2087 to be swapped out again because it is already in the swap area. This
2093 Percentage of cached swap memory in relation to the amount of used swap space.
2096 .IP "-s [ hh:mm[:ss] ]"
2097 Set the starting time of the data, causing the
2099 command to extract records time-tagged at, or following, the time
2100 specified. The default starting time is 08:00:00.
2101 Hours must be given in 24-hour format. This option can be
2102 used only when data are read from a file (option -f).
2104 Indicate which data collector is called by
2106 If the data collector is sought in PATH then enter "which sadc" to
2107 know where it is located.
2109 When reading data from a daily data file, indicate that
2111 should display the timestamps in the original local time of
2112 the data file creator. Without this option, the
2114 command displays the timestamps in the user's locale time.
2116 Report CPU utilization. The
2118 keyword indicates that all the CPU fields should be displayed.
2119 The report may show the following fields:
2124 Percentage of CPU utilization that occurred while executing at the user
2125 level (application). Note that this field includes time spent running
2131 Percentage of CPU utilization that occurred while executing at the user
2132 level (application). Note that this field does NOT include time spent
2133 running virtual processors.
2138 Percentage of CPU utilization that occurred while executing at the user
2139 level with nice priority.
2144 Percentage of CPU utilization that occurred while executing at the system
2145 level (kernel). Note that this field includes time spent servicing
2146 hardware and software interrupts.
2151 Percentage of CPU utilization that occurred while executing at the system
2152 level (kernel). Note that this field does NOT include time spent servicing
2153 hardware or software interrupts.
2158 Percentage of time that the CPU or CPUs were idle during which
2159 the system had an outstanding disk I/O request.
2164 Percentage of time spent in involuntary wait by the virtual CPU
2165 or CPUs while the hypervisor was servicing another virtual processor.
2170 Percentage of time spent by the CPU or CPUs to service hardware interrupts.
2175 Percentage of time spent by the CPU or CPUs to service software interrupts.
2180 Percentage of time spent by the CPU or CPUs to run a virtual processor.
2185 Percentage of time spent by the CPU or CPUs to run a niced guest.
2190 Percentage of time that the CPU or CPUs were idle and the system
2191 did not have an outstanding disk I/O request.
2195 Print version number then exit.
2197 Report status of inode, file and other kernel tables.
2198 The following values are displayed:
2203 Number of unused cache entries in the directory cache.
2208 Number of file handles used by the system.
2213 Number of inode handlers used by the system.
2218 Number of pseudo-terminals used by the system.
2222 Report swapping statistics. The following values are displayed:
2227 Total number of swap pages the system brought in per second.
2232 Total number of swap pages the system brought out per second.
2236 Report task creation and system switching activity.
2241 Total number of tasks created per second.
2246 Total number of context switches per second.
2250 Report TTY devices activity. The following values are displayed:
2255 Number of receive interrupts per second for current serial line. Serial line number
2256 is given in the TTY column.
2261 Number of transmit interrupts per second for current serial line.
2266 Number of frame errors per second for current serial line.
2271 Number of parity errors per second for current serial line.
2276 Number of breaks per second for current serial line.
2281 Number of overrun errors per second for current serial line.
2287 to omit output for any devices for which there was no activity during the
2293 command takes into account the following environment variables:
2296 When this variable is set, display statistics in color on the terminal.
2297 Possible values for this variable are
2302 (the latter is the default).
2304 Please note that the color (being red, yellow, or some other color) used to display a value
2305 is not indicative of any kind of issue simply because of the color. It only indicates different
2309 Specify the colors and other attributes used to display statistics on the terminal.
2310 Its value is a colon-separated list of capabilities that defaults to
2311 .BR C=33;22:H=31;1:I=32;22:M=35;1:N=34;1:R=31;22:Z=34;22 .
2312 Supported capabilities are:
2317 SGR (Select Graphic Rendition) substring for comments inserted in the binary daily
2322 SGR substring for percentage values greater than or equal to 75%.
2326 SGR substring for item names or values (eg. network interfaces, CPU number...)
2330 SGR substring for percentage values in the range from 50% to 75%.
2334 SGR substring for non-zero statistics values.
2338 SGR substring for restart messages.
2342 SGR substring for zero values.
2346 If this variable exists and its value is
2350 will save its data in UTC time (data will still be displayed in local time).
2352 will also use UTC time instead of local time to determine the current daily
2353 data file located in the
2355 directory. This variable may be useful for servers with users located across
2359 If this variable exists and its value is
2361 then the current locale will be ignored when printing the date in the report header.
2364 command will use the ISO 8601 format (YYYY-MM-DD) instead.
2365 The timestamp will also be compliant with ISO 8601 format.
2369 Report CPU utilization for each 2 seconds. 5 lines are displayed.
2372 .B sar -I 14 -o int14.file 2 10
2374 Report statistics on IRQ 14 for each 2 seconds. 10 lines are displayed.
2375 Data are stored in a file called
2379 .B sar -r -n DEV -f @SA_DIR@/sa16
2381 Display memory and network statistics saved in daily data file 'sa16'.
2386 Display all the statistics saved in current daily data file.
2389 filesystem must be mounted for the
2393 All the statistics are not necessarily available, depending on the kernel version used.
2395 assumes that you are using at least a 2.6 kernel.
2399 .I @SA_DIR@/saYYYYMMDD
2401 The standard system activity daily data files and their default location.
2402 YYYY stands for the current year, MM for the current month and DD for the
2409 contain various files with system statistics.
2411 Sebastien Godard (sysstat <at> orange.fr)
2423 .I http://pagesperso-orange.fr/sebastien.godard/