Linux network errors when connecting to datacenter?

Question

linux

Hello dear ones. There was a problem connecting the server in the data center.
OS Linux 3.2.0-4-amd64 #1 SMP Debian 3.2.63-2+deb7u1 x86_64
now there is running a server on CentOS but it needs to be replaced. IP addresses are allocated static. ip 178.222.80.34 mask 255.255.255.252 getwey 178.222.80.33
/etc/network/interfaces
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).
# The loopback network interface
auto lo
auto eth0
iface lo inet loopback #Main
interface
allow-hotplug eth0
iface eth0 inet static
address 178.222.80.34
netmask 255.255.255.252
gateway 178.222.80.33
ip pings in the data center, localhost too
The whole paradox is that everything works fine at home. :o To check in the router, we assigned the geteway IP and the mask indicated above. And disabled DHCP for fidelity. When the server was connected to the network with this configuration, the server immediately connected to the Internet as expected and everything worked. When connecting to the data center with the same settings, an error pops up on top of the console and the server cannot connect to the Internet. What could be the problem? We connect the old server in place, everything connects and works
, here are the lines from the error log:
Dec 6 03:16:16 1000ukg kernel: [ 25.930706] e1000e 0000:03:00.0: eth1: Reset adapter unexpectedly
Dec 6 03:16:20 1000ukg kernel: [ 29.210375] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
Dec 6 03:16:24 1000ukg kernel: [ 33.928163 ] e1000e 0000:03:00.0: eth1: Reset adapter unexpectedly
Dec 6 03:16:28 1000ukg kernel: [ 37.175837] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
Dec 6 03:16:32 1000ukg kernel: [ 41.925585] e1000e 0000:03:00.0: eth1 : Reset adapter unexpectedly
Dec 6 03:16:36 1000ukg kernel: [ 45.289266] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None Dec 6 03:16:43 1000ukg kernel: [ 53.146770] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None Dec 6 03:16:47 1000ukg kernel: [ 56.920793] e1000e 0000:03:00.0: eth1: Reset adapter unexpectedly Dec 6 03:16:50 1000ukg kernel: [ 60.092592] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
Dec 6 03:16:40 1000ukg kernel: [ 49.923040] e1000e 0000:03:00.0: eth1: Reset adapter unexpectedly
Dec 6 03:16:41 1000ukg ntpdate[1929]: Can't find host 0.debian.pool.ntp.org: Name or service not known (-2)
Dec 6 03:16:54 1000ukg kernel: [ 63.918601] e1000e 0000:03:00.0: eth1: Reset adapter unexpectedly
Dec 6 03:16:58 1000ukg kernel: [ 67.190304] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
An attempt to register routes with the route command also gave nothing. ifconf showed that everything is normal. If necessary, I will simulate the situation at home and show. Tried to connect the server twice already, same thing. Everything works at home.
Already broke his head. Help me to understand. What input is needed to resolve this issue? I will provide.
Problem solved! This is what was done, and which of this helped I can’t say for sure, there was no time in the DC for verification!
1. Correct the line GRUB_CMDLINE_LINUX_DEFAULT="quiet pcie_aspm=off" in the /etc/default/grub file
by adding pcie_aspm=off separated by a space
2. Run update-grub
All operations naturally from root
3. Add to /etc/rc.local
To make changes you can restart sh /etc/rc.local
#disable pause autonegotiate
#/sbin/ethtool -A eth0 autoneg off
#/sbin/ethtool -s eth0 autoneg off
/sbin/ethtool -A eth1 autoneg off
#/sbin/ethtool -s eth1 autoneg off
/sbin/ethtool -s eth1 speed 100 duplex full autoneg off # # Disable automatic speed and duplex detection, set speed
#change tx ring buffer
#/sbin/ethtool -G eth0 tx 4096 #maybe too large(consider 512 ). To increase inte$
#interrupts per second>
/sbin/ethtool -G eth1 tx 4096 #maybe too large(consider 512). To increase inter$
#interrupts per second>
#change rx ring buffer
#/sbin/ethtool -G eth0 rx 128
/sbin/ethtool -G eth1 rx 128
#disable wake on line
#/sbin/ethtool -s eth0 wol d
/sbin/ethtool -s eth1 wol d
#turn off offload
#/sbin/ethtool -K eth0 tx off rx off sg off tso off gso off gro off
/sbin/ethtool -K eth1 tx off rx off sg off tso off gso off gro off
#enable TX pause
#/sbin/ethtool -A eth0 tx on
/sbin/ethtool -A eth1 tx on
#disable ASPM
#/sbin/setpci -s 02:00.0 CAP_EXP+ 10.b=40
#/sbin/setpci -s 00:19.0 CAP_EXP+10.b=40
If you have multiple adapters, then add similar commands for other ethXs
Thank you all for your help!

Answer 1

The problem is this:
Dec 6 03:16:50 1000ukg kernel: [ 60.092592] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
Dec 6 03:16:54 1000ukg kernel: [ 63.918601] e1000e 0000:03:00.0: eth1 : Reset adapter unexpectedly
Dec 6 03:16:58 1000ukg kernel: [ 67.190304] e1000e: eth1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
1) Try updating your drivers
2) Try a different network card.

Answer 2

in the eth1 log, and in the eth0 config,
if
everything miraculously works after that, add allow-hotplug eth0 to the config, and if there is allow-hotplug eth1 auto eth1 , it is possible that
flowcontrol is enabled on the switch, if so, ask to disable

Answer 3

Good day!
Does "Reset adapter unexpectedly" bother you?
As far as I remember, it's a hardware problem. I recommend updating your drivers.
On account of the fact that "works at home, does not work in the data center", perhaps due to the type of connection, and the port setting on the switch from the data center

Answer 4

Update the OS kernel.
What kind of motherboard and network?