NIC speed 100mbps instead of 1000mbps

[rcvassallo]

I have two HP DL380G6 servers running ESX4i. These servers are equipped with four 1-gigabit NICs. On Server1 all 4NICs are configured and running at 1000mbps.

However, Server2 indicates that one of the network cards is working at 100mbps instead of 1000mbps. 100mbps is listed for both the configured and actual speed, but I can't seem to change it within vSphere/vCenter. The other 3 NICs in server2 are running at 1000mbps.

Any advice would be greatly appreciated!

[MattG]

Your physical switch port could be misconfigured or your patch cable could be bad.

To attempt to force it to a higher speed in vSphere Client:

Configuration -> Networking -> Click Properties of vSwitch pNic is associated with-> Network Adpaters-> Click in question pNIC->Click Edit button->Change Speed/Duplex setting.

-MattG

If you find this information useful, please award points for "correct" or "helpful".

[rcvassallo]

All the physical ports on my switch are set to auto-sense the link speed. I have tried 3 different ports with the same result. The other 7 NICs are connected to this same switch without any issues.

When I attempted to force it to run at 1000mbps, I received the following error:

Error during the configuration of the host. Forcing 1000mbps or faster may not be supported on this network device.

[MattG]

Have you tried changing patch cables?

[rcvassallo]

The patch cables are all brand new Cat5e, but I went ahead and swapped it out anyway. vCenter still shows 100mbps.

The switch is HP Procurve 1810. The management interface reports that the NIC-in-question running at 100mbps half duplex mode.

I looked around in the BIOS of the host DL380G6 I did not find any settings related to the NIC speed.

Appreciate the help. Any other ideas?

[Damon123]

I am experiencing the exact same issue. My nic speed keeps auto negotiating to 100mbps even though it's a 1000mbps card. The switch is also a gb switch. The port is configured exactly the same as all the other ports on the switch. The hardware (HP DL380 G5) is exactly the same hardware as the other vmware hosts plugged into the switch and they're all autonegotiating to 1000mbps. The only difference between those machine and this one is that my other machines are running esxi 3.5 and esx 4.0. This machine, the one I'm having trouble with is running esxi 4.0 update 1 - the HP Proliant specific one.

[rcvassallo]

Well at least I'm not the only one with this issue.

Both of my DL380G6 hosts are running ESXi 4.0.0 release 208167 which I believe is the Update 1 build.

What has me most confused is that my 2 hosts are basically identical. The server hardware is the same and all NICs are connected to the same switch, yet only one of the NICs comes up as 100mbps

[Damon123]

Ok, so in my case it actually was an issue with the port itself. My issue is now resolved. Even though I'd replaced all the cables and tried plugging into various ports on my switch, the bad port was actually in our patch panel!! I'm using a different port in the patch panel now and it autonegotiated to 1000 mbps full.

[rcvassallo]

Unfortunately for me there is no patch panel in the mix. The servers are connected directly to the switch. I've tried 3 different ports with the same result. I will have to see if power-cycling the switch does any good but other than that I am at a loss for resolving this.

[a_p_]

If you are able to connect to ILO2 management using the dedicated ILO2 port then you don't need to read any further.

If it is NIC1 which runs on 100 Mbit/s, you could check, if for any reason ILO2 is configured for network port sharing.

To check this, you have to reboot the server:

- When prompted during POST, press the F8 key to enter iLO 2 RBSU.

- Select Network > NIC > TCP/IP, and press the Enter key.

- "Network Interface Adapter" should be set to "ON" (NOT "Shared Network Port")

Actually this should not have anything to do with the speed, however who knows!?

(see "HP Integrated Lights-Out 2 User Guide" for details)

André

[mdwasim]

Anyone got solution for NIC speed?

Same issue with Broadcom 5709 on Dell R710 server.

This issue occurs only after upgrading ESX 4.0 to Update 2.

This shows the NIC speed

This shows the upload speed before updating to update 2

[kpate96]

Did anyone ever figure this out? I am having the same issue with a brand new Cisco UCS C200M2 running ESXi 4.1.

I have tried changing the speed in the Vsphere Client and CLI and both give me the same error.

Also when i manually set the speed on the switch it only works if I set it to 10 or 100 Mbps...

This is very strange since the NIC is a gigabit NIC.

[NTShad0w]

there are some reasons of 100Mbps NIC connection (especially on HP DL380 G5) on Vmware ESX/ESXi v4 and v5

1. HP DL380 G5 have ILO NIC on the first NIC so when it's in shared mode it's act as a 100Mbps or eventually 10Mbps NIC

2. ALWAYS check the cable, its really often happen that the cable is a real problem, of course patch panell too

3. server/NIC firmware problem, it's happens but not so often

4. broken NIC, it's hapen rather rare

5. bad negociation on switches and NIC... it's happens sometimes

regards

NTShad0w

[jhollinger]

I had this problem also.
Config: 2 Dell R610 servers each with 4 x 1G ports.

The problem was ONE port on ONE server was reporting 100Mbps... I unplugged, replugged. I swapped cables. Nothing changed. Only when I changed ports on the switch did it make a difference. I am still not sure why. I need to test another 1Gbps device on that port.

[BobCoble]

I was having the same issue.  I'm using a DELL 610 server w/ 4-1GB NICs connected to a switch equipped with a 1GB RJ-45 SFP.  Turned out to be a problem w/ the SFP - replaced the Methode (don't know part #) with a Finisar FCMJ-8521-3 and all good now.

[mbutler101]

FYI, my issue turned out to be a bad cable...switched the patch cable and was able to achieve 1000mbps.

I thought it might be an issue with the Dell iDRAC card on this Dell R610 server sharing the same network port, since the iDRAC has a max port speed of 100mbps, but I eliminated that as the source of the issue by switching this iDRAC to use a dedicated port (this was an iDRAC enterprise card).

[jionx]

I met the same problem today. My server is a new DELL R530 just bought this month (Feb 2017).  One of the port always shown as 100M.

It finally turned out to be the bad network cable issue.  After I replaced the cable, it is normal now and running at 1000M.

[Brandon_Buck]

Ditto,

I'm only replying to this post to reaffirm to someone else with same issue as well. Same issue 100mb no matter what switch port, turned out to be a bad cable.

[fireskyer]

HP DL 380 G5 ( Updated with latest Supported SPP  HP_Service_Pack_for_Proliant_2014.06.)

ESXI 5.5 : 5.5 5230635

Dito with an similar Problem. We set our NIC Adapter  to set the Speed manually then comes an error so we had to switch back to negotiation .. and he assigns us 100 mbit. 

How i can check the negotiation from esxi to the switch ?

Don't know where the Problem lies.

greetings  john

[samkat2008]

I agree with your assertion. I swapped an locally-made Cat6 cable for a factory made Cat6E cable and the ports automatically moved to 1000Mbps. My set up is a back-to-back connection for vMotion on IBM System X servers.