Shulou(Shulou.com)06/01 Report--

This article shows you how to install and use the system monitoring tool Collectl in Linux. The content is concise and easy to understand, which will definitely brighten your eyes. I hope you can get something through the detailed introduction of this article.

System resource monitoring

As a lightweight monitoring tool, Collectl is the most functional tool of its kind. Users can monitor different complex system matrix values and retain data for later analysis. Unlike other tools that only monitor specific system parameters, Collectl can monitor different variables at the same time and record them in an appropriate manner.

Unlike or focusing on a small group of statistics, using a unique output mode, or using iteration as most of the monitoring tools that the daemon runs, collectl can all be implemented at the same time. Users can select any of various subsystems to monitor information including memory, CPU, disk, Inode, wireless bandwidth, lustre, memory, network, network file system, process, quadratic, slabs, socket and TCP.

Let's take a look at the following commands before we go any further.

$collectl

Waiting for1 second sample. . .

# "- CPU-"- Disks-"- Network-"

# cpu sys inter ctxsw KBRead Reads KBWrit Writes KBIn PktIn KBOut PktOut

00864177200000100

521338273400820001

1012222647009230201

107631722008030102

CPU usage, disk input and output, and network activity are loaded per second. The information is easy for people who can understand the data. This list continues to grow over a given interval and can directly form a file tracking log. The collectl tool provides a variety of commands to record, find, and do other processing of this data.

Install collectl tools

Collectl is the default resource for Ubuntu/Debian users, so you can get it using the apt command.

$sudo apt-get install collectl

Fedora/CentOS users can get it by using the yum command.

$yum install collectl

Use

Necessary understanding-Collectl subsystem

Subsystems are different types of system resources that can be detected. Such as CPU, memory, bandwidth and so on can form a subsystem. Running only the collectl command will output CPU, disk, and network subsystem information in batch mode, as we saw above.

As you can see from the instructions, Collectl can identify the following subsystems.

Total subsystem

B-buddy info (memory fragment)

C-CPU

D-Disk

F-NFS V3 Data

I-InodeandFileSystem

J-Interrupts

L-Lustre

M-Memory

N-Networks

S-Sockets

T-TCP

X-Interconnect

Y-Slabs (system object caches)

Detail subsystem

This is a series of details, which in most cases are derived from the corresponding total data. At present, there is no corresponding total data for the two categories "environment variables" and "processes". If you have three disks, select-sd, and you will only see a single message from a combination of three disks. If you select-sD, each disk information will be displayed separately.

C-CPU

D-Disk

E-Environmental data (fan, power, temp), via ipmitool

F-NFS Data

J-Interrupts

L-Lustre OST detail OR client Filesystem detail

M-Memory node data, which is also known as numa data

N-Networks

T-65 TCP counters only available in plot format

X-Interconnect

Y-Slabs (system object caches)

Z-Processes

Use "- s" to monitor a specific subsystem and add subsystem identifiers to it. Now let's give a few examples.

1. Monitor cpu utilization

Using "- sc" to monitor the total utilization rate of CPU

$collectl-sc

Waiting for1 second sample. . .

# "- CPU-"

# cpu sys inter ctxsw

3018003729

3017673599

Use "- C" to observe individual CPU usage. The result will be multiple lines, each corresponding to a CPU.

$collectl-sC

Waiting for1 second sample. . .

# SINGLE CPU STATISTICS

# Cpu User Nice Sys Wait IRQ Soft Steal Idle

0300000096

1300000096

2200000097

3100000098

0200000097

1202000095

2100000098

3401000095

If necessary, you can use both C and c to obtain individual CPU monitoring data and summary data.

two。 Memory monitoring

Use the m subsystem to view memory.

$collectl-sm

Waiting for1 second sample. . .

# "- Memory-"

# Free Buff Cach Inac Slab Map

2G220M1G1G210M3G

2G220M1G1G210M3G

2G220M1G1G210M3G

It's not hard to explain. M is used to view more memory details.

$collectl-sM

Waiting for1 second sample. . .

# MEMORY STATISTICS

# Node Total Used Free Slab Mapped Anon Locked Inact Hit%

07975M5939M2036M215720K372184K06652K1434M0

07975M5939M2036M215720K372072K06652K1433M0

3. View disk usage

D and D can view an overview and details of disk usage.

$collectl-sd

Waiting for1 second sample. . .

# "- Disks-"

# KBRead Reads KBWrit Writes

4113624

008013

$collectl-sD

Waiting for1 second sample. . .

# DISK STATISTICS (/ sec)

# "- reads-"- writes-"- averages-" Pct

# Name KBytes Merged IOs Size KBytes Merged IOs Size RWSize QLen Wait SvcTim Util

Sda 0000000000000

Sda 0000000000000

Sda 10211715322621

Sda 0000921151818112125

The "--verbose" command can be used to view additional information. Unlike the D command, it extends the profile to include more information.

$collectl-sd-verbose

4. Report on multiple systems at the same time

If you want to get CPU, memory and disk reports at the same time, use a combination of subcommands at the same time.

$collectl-scmd

Waiting for1 second sample. . .

# "- CPU-"- Memory-"- Disks-"

# cpu sys inter ctxsw Free Buff Cach Inac Slab Map KBRead Reads KBWrit Writes

40218743341G221M1G1G210M3G0000

30189640651G221M1G1G210M3G00205

5. Show statistical time

To display the monitoring information for each line with the time, use the T option. It is specified by the "- o" parameter.

$collectl-scmd-oT

Waiting for1 second sample. . .

# "- CPU-"- Memory-"- Disks-"

# Time cpu sys inter ctxsw Free Buff Cach Inac Slab Map KBRead Reads KBWrit Writes

12:03:0530196140131G225M1G1G212M3G0000

12:03:0630188438101G225M1G1G212M3G0000

12:03:0730201140601G225M1G1G212M3G0000

Use "- oTm" to change the time to millisecond display.

6. Change the sample count

Each line of the collectl report is a snapshot or sample, and it usually sets periodic intervals such as 1 second to sample. I can be used to set the time interval and c to set the sample count.

$collectl-C1-sm

Waiting for1 second sample. . .

# "- Memory-"

# Free Buff Cach Inac Slab Map

1G261M1G1G228M3G

Use the I command to change the update interval.

$collectl-sm-i2

Waiting for2 second sample. . .

# "- Memory-"

# Free Buff Cach Inac Slab Map

1G261M1G1G229M3G

The above command sets memory information to be collected every 2 seconds.

7. Use collectl like iotop

The top command makes collectl do intelligent statistics like an iostat/top tool. The list is constantly updated and can be sorted using different fields.

$collectl-top iokb

The output is as follows

# TOP PROCESSES sorted by iokb (counters are / sec) 09:44:57

# PID User PR PPID THRD S VSZ RSS CP SysT UsrT Pct AccuTime RKB WKB MajF MinF Command

3104 enlighte 2026833 S 938M33M00.000.00000:09.160400/usr/bin/ktorrent

1 root 2000 S 26M3M20.000.00000:01.300000/sbin/init

2 root 2000 S 0030.000.00000:00.000000 kthreadd

3 root 2020 S 0000.000.00000:00.020000 ksoftirqd/0

4 root 2020 S 0000.000.00000:00.000000 kworker/0:0

5 root 020 S 0000.000.00000:00.000000 kworker/0:0H

7 root RT 20 S 0000.000.00000:00.080000 migration/0

8 root 2020 S 0020.000.00000:00.000000 rcu_bh

9 root 2020 S 0000.000.00000:00.000000 rcuob/0

The above output is very similar to the top command, and it sorts processes in descending order of the number of disks.

If you want to show only the above five processes, you can use the following command

$collectl-top iokb,5

To learn which fields in the list above can be sorted, use the following command

$collectl-showtopopts

The following is a list of the highest sort types applied to processes or data. In some cases you may use a field to sort, but it is not part of the display.

TOP PROCESS SORT FIELDS

Process sort field

Memory

Vsz virtual memory

Rss resident (physical) memory

Time

Syst system time

Usrt user time

Time total time

Accum accumulated time

I/O

Rkb KB read

Wkb KB written

Iokb total I/O KB

Rkbc KB read from pagecache

Wkbc KB written to pagecache

Iokbc total pagecacge I/O

Ioall total I take O KB (iokb+iokbc)

Rsys read system calls

Wsys write system calls

Iosys total system calls

Iocncl Cancelled write bytes

PageFaults

Majf major page faults

Minf minor page faults

Flt total page faults

ContextSwitches

Vctx volunary context switches

Nctx non-voluntary context switches

Miscellaneous (best when used with--procfilt)

Cpu cpu number

Pid process pid

Thread total process threads (not counting main)

TOP SLAB SORT FIELDS

Numobj total number of slab objects

Actobj active slab objects

Objsize sizes of slab objects

Numslab number of slabs

Objslab number of objects in a slab

Totsize total memory sizes taken by slabs

Totchg change in memory sizes

Totpct percent change in memory sizes

Name slab names

8. Use collectl like top

To make collectl look like top, we just need to sort the output processes by CPU usage.

$collectl-top

The output is as follows

# TOP PROCESSES sorted by time (counters are / sec) 14:08:46

# PID User PR PPID THRD S VSZ RSS CP SysT UsrT Pct AccuTime RKB WKB MajF MinF Command

9471 enlighte 2091020 R 63M22M30.030.101300:00.810003/usr/bin/perl

3076 enlighte 2026832 S 521M40M20.000.03300:55.140002/usr/bin/yakuake

3877 enlighte 20335641 S 1G218M10.000.03310:10.500000/opt/google/chrome/chrome

4625 enlighte 20289536 S 1G241M20.000.02208:24.3900012/usr/lib/firefox/firefox

5638 enlighte 2033563 S 1G265M10.000.02209:55.040002/opt/google/chrome/chrome

1186 root 2011524 S 502M76M00.000.01103:02.960000/usr/bin/X

1334 www-data 2013290 S 87M1M20.000.01100:00.850000 nginx:

The above command can also be used to display subsystem information.

$collectl-top-scm

9. List processes like ps

To list all processes like the ps command with no subsequent updates, use the "c" command to count them to 1. $collectl-C1-sZ-iV1

The above command will list all processes similar to the "ps-e" command. "procfilt" is used to filter out specific process information from all processes. "procopts" is used to specify another set of commands to fine-tune the list of processes.

10. Use collectl like vmstat

Collectl has built-in commands to perform functions like vmstat.

$collectl-vmstat

Waiting for1 second sample. . .

# procs-memory (KB)-swaps---io-----system-- cpu-

# r b swpd free buff cache inact active si so bi bo in cs us sy id wa

1001733M242M1922M1137M710M0001081982391820951

1001733M242M1922M1137M710M00001906388610980

1001733M242M1922M1137M710M00001739348030960

11. Details of the subsystem

The following command counts CPU information five times at one-second intervals and displays detailed information along with the time (verbose).

$collectl-sc-c5-i1-- verbose-oT

Waiting for1 second sample. . .

# CPU SUMMARY (INTR, CTXSW & PROC / sec)

# Time User Nice Sys Wait IRQ Soft Steal Idle CPUs Intr Ctxsw Proc RunQ Run Avg1 Avg5 Avg15 RunT BlkT

14:22:101100000087413122691086610.780.860.7810

14:22:111500000084412832496086610.780.860.7810

14:22:121700000082413422658086600.780.860.7800

14:22:131500000084412412429086610.780.860.7810

14:22:141100000088412702488086600.800.870.7800

Change the "- s" variable to see the different subsystems.

The above is how to install and use the system monitoring tool Collectl for Linux. Have you learned any knowledge or skills? If you want to learn more skills or enrich your knowledge reserve, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.