Network tools

2024-04-19 08:42:50  Updated

sar

You can use sar to obtain system activity reports, which show the file read/write status, system call usage, disk I/O, CPU utilization, memory usage, process activities, and IPC-related activities. You can use sar to monitor network conditions.

Synopsis:

sar -n [keyword]  [ <interval> [ <count> ] ]

The parameters are described as follows:

  • -n: monitors network performance.

  • Valid keywords:

    • DEV: displays network interface information.
    • EDEV: displays statistics on network errors.
    • NFS: collects statistics about active NFS clients.
    • NFSD: collects statistics about the NFS server.
    • SOCK: displays socket information.
    • ALL: displays all five switches.
  • interval: the output refresh interval.

  • count: the number of output times.

You can run the sar -n command to view the transmission rate on each NIC. Here is an example:

$sar -n DEV 1
09:01:53 PM     IFACE   rxpck/s   txpck/s    rxkB/s    txkB/s   rxcmp/s   txcmp/s  rxmcst/s
09:01:54 PM        lo      7.07      7.07      0.45      0.45      0.00      0.00      0.00
09:01:54 PM      eth0     62.63     19.19      4.18      2.57      0.00      0.00      0.00
09:01:54 PM      eth1      0.00      0.00      0.00      0.00      0.00      0.00      0.00
09:01:54 PM     bond0     62.63     19.19      4.18      2.57      0.00      0.00      0.00

The fields are described as follows:

  • IFACE: the name of the server NIC.

  • rxpck/s: the number of packets received per second.

  • txpck/s: the number of packets sent per second.

  • rxKB/S: the size of packets received per second, in KB.

  • txKB/S: the size of packets sent per second, in KB.

  • rxcmp/s: the number of compressed packets received per second.

  • txcmp/s: the number of compressed packets sent per second.

  • rxmcst/s: the number of multicast packets received per second.

vsar

You can use vsar to trace the network load over a period.

Run the vsar --traffic --tcp command to trace the historical network load. Here is an example:

Time              -------------------------------tcp------------------------------ ---------------------traffic--------------------
Time              active  pasive    iseg  outseg  EstRes  AtmpFa  CurrEs  retran    bytin  bytout   pktin  pktout  pkterr  pktdrp
18/10/21-16:22:28   2.20   27.40  249.60  222.20    0.20    0.00  223.00    0.00    50.0K   37.2K  249.00  202.00    0.00    0.00
18/10/21-16:22:33   3.20   20.60  247.40  230.60    0.00    0.00  198.00    0.00    76.6K   32.0K  256.00  205.00    0.00    0.00
18/10/21-16:22:38   2.00   25.60    3.6K   50.5K    0.00    0.00  216.00    0.01   276.8K   73.4M    3.6K   50.5K    0.00    0.00
18/10/21-16:22:43   3.80   29.00  243.80  214.60    0.00    0.00  232.00    0.47    53.1K   32.1K  255.00  206.00    0.00    0.00
18/10/21-16:22:48   2.00   31.20  251.20  217.00    0.00    1.00  238.00    0.46    53.5K   35.1K  274.00  221.00    0.00    0.00
18/10/21-16:22:53   1.20   25.60    2.6K   37.1K    0.00    0.00  228.00    0.00   207.8K   54.1M    2.6K   37.1K    0.00    0.00
18/10/21-16:22:58   0.40   27.00    1.2K   16.6K    0.00    0.00  225.00    0.00   110.0K   24.1M    1.2K   16.6K    0.00    0.00
18/10/21-16:23:03   2.40   25.60    3.7K   50.4K    0.00    0.00  217.00    0.00   281.3K   73.4M    3.7K   50.4K    0.00    0.00

The fields are described as follows:

  • tcp

    • active: the number of connections initiated actively per second.
    • pasive: the number of passively established connections per second.
    • iseg: the number of inbound TCP packets per second.
    • outseg: the number of outbound TCP packets per second.
    • EstRes: the number of connections reset per second.
    • AtmpFa: the number of connection initiation failures per second. For example, the failure can be caused if the syn packet is discarded because the semi-join queue is full.
    • CurrEs: the number of current TCP connections.
    • retran: the retransmission rate calculated based on the retransmitted segments and the total segments.
  • traffic

    • bytin: the number of bytes received by the physical host NIC per second. Generally, 960 MB is the upper limit for a 10 GE NIC.
    • bytout: the number of bytes sent by the physical host NIC per second.
    • pktin: the number of packets received by the physical host NIC per second.
    • pktout: the number of packets sent by the physical host NIC per second.
    • inerr: the number of inbound packet errors.
    • outerr: the number of outbound packet errors.
    • indrp: the number of inbound packets dropped.
    • outdrp: the number of outbound packets dropped.

Contact Us