R
R
Rafail Bektemirov2019-07-08 01:34:05
Debian
Rafail Bektemirov, 2019-07-08 01:34:05

Increase network connection speed to 1Gb/s?

Good afternoon,
There is a server (intel s2600cp + 2 * e5-2620) with 4 network cards on board, PROXMOX virtualization system.
Switch 24 port, those. all ports are gigabit with a bandwidth of 48Gb / s
While 1 network card is configured and used. Bridge tuned.

spoiler

Линк на все сетевухи 1Гб
[email protected]:~# ip link list
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: enp9s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq master vmbr0 state UP mode DEFAULT group default qlen 1000
    link/ether 00:1e:67:b7:74:10 brd ff:ff:ff:ff:ff:ff
3: enp9s0f1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq master vmbr1 state UP mode DEFAULT group default qlen 1000
    link/ether 00:1e:67:b7:74:11 brd ff:ff:ff:ff:ff:ff
4: enp9s0f2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether 00:1e:67:b7:74:12 brd ff:ff:ff:ff:ff:ff
5: enp9s0f3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
    link/ether 00:1e:67:b7:74:13 brd ff:ff:ff:ff:ff:ff
6: vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
    link/ether 00:1e:67:b7:74:10 brd ff:ff:ff:ff:ff:ff
7: vmbr1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
    link/ether 00:1e:67:b7:74:11 brd ff:ff:ff:ff:ff:ff
8: tap100i0: <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master vmbr0 state UNKNOWN mode DEFAULT group default qlen 1000

iprerf3 - между виртуальными серверами и хостом примерно одинаковый +/-
[email protected]:~# iperf3 -c 192.168.1.250
Connecting to host 192.168.1.250, port 5201
[  4] local 192.168.1.254 port 51142 connected to 192.168.1.250 port 5201
[ ID] Interval           Transfer     Bandwidth       Retr  Cwnd
[  4]   0.00-1.00   sec  1010 MBytes  8.47 Gbits/sec    0    437 KBytes
[  4]   1.00-2.00   sec   941 MBytes  7.89 Gbits/sec    0    437 KBytes
[  4]   2.00-3.00   sec   955 MBytes  8.01 Gbits/sec    0    437 KBytes
[  4]   3.00-4.00   sec   969 MBytes  8.13 Gbits/sec    0    437 KBytes
[  4]   4.00-5.00   sec  1.02 GBytes  8.72 Gbits/sec    0    437 KBytes
[  4]   5.00-6.00   sec  1.03 GBytes  8.86 Gbits/sec    0    437 KBytes
[  4]   6.00-7.00   sec   985 MBytes  8.26 Gbits/sec    0    437 KBytes
[  4]   7.00-8.00   sec   801 MBytes  6.72 Gbits/sec    0    437 KBytes
[  4]   8.00-9.00   sec  1.04 GBytes  8.94 Gbits/sec    0    437 KBytes
[  4]   9.00-10.00  sec  1.06 GBytes  9.09 Gbits/sec    0    437 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-10.00  sec  9.67 GBytes  8.31 Gbits/sec    0             sender
[  4]   0.00-10.00  sec  9.67 GBytes  8.31 Gbits/sec                  receiver

iprerf3 - на комп в сети ч\з свич
[email protected]:~# iperf3 -c 192.168.1.55
Connecting to host 192.168.1.55, port 5201
[  4] local 192.168.1.254 port 34622 connected to 192.168.1.55 port 5201
[ ID] Interval           Transfer     Bandwidth       Retr  Cwnd
[  4]   0.00-1.00   sec  12.1 MBytes   102 Mbits/sec    0    154 KBytes
[  4]   1.00-2.00   sec  11.2 MBytes  94.4 Mbits/sec    0    154 KBytes
[  4]   2.00-3.00   sec  11.2 MBytes  93.8 Mbits/sec    0    154 KBytes
[  4]   3.00-4.00   sec  11.2 MBytes  94.4 Mbits/sec    0    154 KBytes
[  4]   4.00-5.00   sec  11.2 MBytes  93.8 Mbits/sec    0    154 KBytes
[  4]   5.00-6.00   sec  11.2 MBytes  94.4 Mbits/sec    0    154 KBytes
[  4]   6.00-7.00   sec  11.2 MBytes  93.8 Mbits/sec    0    154 KBytes
[  4]   7.00-8.00   sec  11.2 MBytes  94.3 Mbits/sec    0    154 KBytes
[  4]   8.00-9.00   sec  11.2 MBytes  94.4 Mbits/sec    0    154 KBytes
[  4]   9.00-10.00  sec  11.2 MBytes  93.8 Mbits/sec    0    154 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-10.00  sec   113 MBytes  94.9 Mbits/sec    0             sender
[  4]   0.00-10.00  sec   112 MBytes  94.2 Mbits/sec                  receiver


Question 1: Do I understand correctly, in order to increase the speed of the network with devices, etc. computers \servers not virtual servers need to configure bonding on the remaining network cards?
Question 2: The most relevant at the moment is how to make / increase the speed to the nominal value of 1Gb / s for connecting through a switch.
Question 3: It is possible to throw one network from a server into a virtual machine, I think this will be a temporary solution.
Thank you.

Answer the question

In order to leave comments, you need to log in

3 answer(s)
A
Alexey Cheremisin, 2019-07-08
@leahch

1) To make bonding, you need a MANAGED switch. Yours is no good. More precisely - still better manageable with support for 802.3ad LACP
2) The speed of 100Mb is not normal, but the whole question is how and what you measure. I recommend mastering iperf and testing the speed with and without a switch (by direct connection of two adapters).
3) There is an ethtool utility, through which you can see at what speed the adapter connected to the switch. Read how you can, for example, here - lexpr.ru/ethtool
4) The case may be in bad or incorrectly crimped cables. At least see point 3.
5) About virtio Already wrote to you.
6) Yes, you can throw a completely physical adapter into a virtual machine (and even make several virtual ones from one adapter!). To do this, the server must be able to VT-d / IOMMU (enabled in the BIOS), and adapters - SR-IOV / VFIO.

R
Rafail Bektemirov, 2019-07-08
@rafail76

Virtual machines already have the latest virtio 171 drivers from redhat.
This is not the question. And the fact is that the speed outside the virtual switch falls sharply and does not give out more than 100 megabits!

P
Puma Thailand, 2019-07-08
@opium

Set up intelligent bonding on the network card, and on the lacp switch, distribute the receiver and transmitter ports over IP, the traffic is relatively evenly distributed if there are not very few clients

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question