I have 2 servers Windows 2008 R2 x64 that I have just installed X520-DA2 cards. The moment I attach the cable the network adapter disappears from the network connections screen and in the device manager it displays a message that 'Windows has stopped this device because it has reported problems.'
The System event log records an error "The driver was unable to load due to an unsupported SFP+ module installed in the adapter." which seems to indicate that I am using an incompatable cable. However I have reviewed the information at http://www.intel.com/support/network/adapter/pro100/sb/CS-030612.htm http://www.intel.com/support/network/adapter/pro100/sb/CS-030612.htm and it indicates that :
Which direct attach cables are compatible with the Intel Ethernet Server Adapter X520 Series?The Intel Ethernet Server Adapter X520 Series and Intel® 82599 10 Gigabit Ethernet Controller-based connections support any SFP+ passive or active limiting direct attach copper cable that complies with the SFF-8431 v4.1 and SFF-8472 v10.4 specifications.
The cable I am using is a Tripp-Lite N280-05M which they list on their website as being SFF-8431 http://www.tripplite.com/en/products/model.cfm?txtModelID=4904 http://www.tripplite.com/en/products/model.cfm?txtModelID=4904
I have found several other posts here that appear to be similar or identical; however, the posters never posted their solution or perhaps they simply gave up.
The other posts are here:
I hope I am not alone on this one! TIA for any help.
after round about 10 months and a few 1000 euros spent, I can now confirm, that the issue with the HP E3500yl switch, the HP 15m DA cable and the x520-DA2 is now fixed by the new driver. Thanks a lot and please continue to analyze and fix problems ....
What adapters and what cables are you using? I am considering just ordering some different cables but at $100 a pop I am reluctant to just start buying more parts, especially considering the parts that I am using should meet the spec according to the manufacturer and the other posters experience of the same cards and cables working together with a different OS...
2 DEll R900's with fresh installas of 2008r2 with SP1 and all the updates.
2 (1 ea) intel X520-DA2 adapters, which after installation and before cable attachment appeared properly.
2 Tripp Lite N280-5M 5 meter SFP+ to SFP+ High Speed Passive Copper 10Gig Cable
Prior to purchase, I verified HCL for the adpater, which staes it works with any cable meeting SFF-8431 v4.1 and SFF-8472 v10.4 specifications.
The cable states that it is MSA SFF-8431 Compliant so everything should work.
The adapter has drivers specific to the operating system used - Server 2008 R2 ( all editions).
There should be no problems....
but thats why we're here instead of working on our networks/servers.
Before the purchase, have to check with the seller that cable is compatible with Intel 10G NIC. I believe it's the different copper cable. If it's not, the Windows Device Manager gives the NIC icon a yellow triangle with question mark. I bought a new one and it works with Intel X520-DA2 and other 10G NIC cards as well (Emulex, Myricom). So two things, 1) it's different cable from what you have 2) ask the seller for an Intel 10G compatible cable. Hope this helps.
Hi motodave785 and others, thanks for using Intel® Ethernet products.
This issue and others like it are interesting to our support team, and we'd like to let you know we are investigating. It's likely we'll need to contact one or more of the posters in this thread via the forum private messages to do some follow-up debugging, and we'll make sure to update the thread with any progress we've made.
The general issue of cable support can be touchy, even when all of the related specifications seem to match. Any aberrations in the detection of the SFP+ module, be it on the part of the NIC or the cable, ends adbuptly without providing much in the way of failure detail. Our first step will be to get these cables into our lab and see exactly what is going on when they get plugged in.
To be continued...
South @ Intel
I will be sending one of my cables to South @ Intel for them to test in their lab and will post back with results. In the meantime, I am going to try a different manufacturer of Passive Cu cable.
A little update on this one, but first a big thanks to motodave for so graciously lending us some cables!
We have reproduced the failure here, that part was easy. Our next step was to take a look at the cable EEPROM to see what was stored in it. The EEPROM in these cables has information that identifies the cable, and our adapter will take a look there to determine what kind of cable is being plugged in (and whether or not it's supported.)
Well, our first attempts to read these particular cables weren't pretty; half the time we don't see an EEPROM at all, and the other half of the time we read a bunch of junk. Ugly as it is though, that's progress.
Our effort now is focused on figuring out why we're not reading legitimate values. We don't have enough information yet to say for sure what the problem is (and I don't want to jinx the folks working on this by guessing when we will) so I'll just say we're on it and when we know more than so will you!
Thanks for the update South. I will be curious to hear of the progress.
After sending you my cables that I was having problems with, I ordered some Cisco cables (SFP-H10GB-CU3M) and they have been working perfectly.
I am running into the same issue with minor differences in setup. The cards themselves are x520-DA NICs; the cables are HP x240 10G SFP+ (P/N JD097B); the OS is VMware ESXi 5.0. When the machine and OS come up with no cables plugged in, the OS shows all NICs. As soon as I plug in one of the cables, an error occurs in the vmkernel.log ("SFP+ module is not supported") and ports disappear. I am working to replace the NICs with x520-DA2 cards since according to VMware, the x520-DA will not work correctly. However, it looks like the cables themselves may be the issue -- at least if previous posters are any indication. I read the same statement as an earlier poster (paraphrased => if the DA cable complies with spec, it's legit) and figured I should be OK. Have there been any updates on known good cables? It sounded like the Cisco cables were known good. I'm just hesitant to start purchasing loads of expensive cables if it isn't clear which brands are supported fully.
If you look at my last reply, you will see that I was in exactly that same state of mind. I purchased some new cables and presto it worked as expected. I included the part # of the cable that I ordered and they were also 3m cables.
Your cables are either faulty or not actually compatible even though the spec says that they do.
Get some new cables and move on.
I doubt that you need new adapters. The adapters have a product code that ends in DA, like E10G42BTDA. However, these are the same adapters that are named DA2 in the brand name string as in Intel® Ethernet Server Adapter X520-DA2. There is not an older DA model in the X520 series that was replaced by a DA2 model. Motodave785 is on the right track here. The problem is most likely with cable incompatibility.
have a simiular problem.
HP E3500yl SPF+ <----2x-----------------------------------------------------------------------------------------2x------> Intel x520-da2 - Exomium S200 SAN box9287BModul / ProCurve / HP X242 SFP+ to SFP+ 15m Direct
Intel 82598EB 10-Gigabit AT CX4 - Exomium S200 !
CX4 <----2x-----------------------------------------------------------------------------------------2x------> 2 vSphere 5 boxes
On the right site is also a Intel Corporation 82598EB 10-Gigabit AT CX4 card installed, which works fine since a couple of month.
The links comes up and works for a couple of hours. After a while the link becomes disconnected, the switch shows the link a "Down".
The right site a OPEN-E (Linux) based NAS/SAN box. Firmware of the switch is up to date.
I get a lot RX errors, no TX error. Driver version: 3.3.9-NAPI.
Since 2 days I get also the stack trace below (ixgbe_alloc_rx_buffers) which would explain the RX errors. I opened a ticket at the vendor of the box,
but may your can point me in the right direction to fix the problem.
Thanks a lot in advance
eth4 Link encap:Ethernet HWaddr 00:1B:21:D4:63:D0
inet addr:172.20.10.20 Bcast:172.20.10.255 Mask:255.255.255.0
UP BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:85853641 errors:106711 dropped:0 overruns:0 frame:106711
TX packets:101422994 errors:0 dropped:0 overruns:0 carrier:0
RX bytes:1002864522968 (933.9 GiB) TX bytes:245777016422 (228.8 GiB)
eth5 Link encap:Ethernet HWaddr 00:1B:21:D4:63:D1
inet addr:172.20.0.20 Bcast:172.20.1.255 Mask:255.255.254.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:4179747 errors:649940 dropped:0 overruns:0 frame:649940
TX packets:4662351 errors:0 dropped:0 overruns:0 carrier:0
RX bytes:2049773029 (1.9 GiB) TX bytes:3112619034 (2.8 GiB)
author: Intel Corporation, <<a href="mailto:email@example.com">firstname.lastname@example.org <<a href="mailto:email@example.com">mailto:firstname.lastname@example.org>>
description: Intel(R) 10 Gigabit PCI Express Network Driver
vermagic: 188.8.131.52-oe64-00000-g5ec1914 SMP mod_unload2011/12/04 00:20:04|[379834.068214] Pid: 18059, comm: rrd_pysnmp_grap Not tainted 184.108.40.206-oe64-00000-g5ec1914 # 492011/12/04 00:20:04|[379834.068217] 2011/12/04 00:20:04|[379834.068217] Call Trace:2011/12/04 00:20:04|[379834.068219]  __alloc_pages_internal+0x3fd/0x4502011/12/04 00:20:04|[379834.068227]  cache_alloc_refill+0x2c3/0x5e02011/12/04 00:20:04|[379834.068230]  __kmalloc+0xfc/0x1102011/12/04 00:20:04|[379834.068233]  __alloc_skb+0x6a/0x1502011/12/04 00:20:04|[379834.068248]  ixgbe_alloc_rx_buffers+0x1ef/0x3b0 [ixgbe]2011/12/04 00:20:04|[379834.068258]  ixgbe_poll+0xbf2/0x1340 [ixgbe]2011/12/04 00:20:04|[379834.068268]  e1000_poll+0x51b/0x5b0 [e1000e]2011/12/04 00:20:04|[379834.068271]  net_rx_action+0x13b/0x2002011/12/04 00:20:04|[379834.068274]  __do_softirq+0x73/0xf02011/12/04 00:20:04|[379834.068277]  call_softirq+0x1c/0x302011/12/04 00:20:04|[379834.068280]  do_softirq+0x35/0x702011/12/04 00:20:04|[379834.068282]  irq_exit+0x89/0x902011/12/04 00:20:04|[379834.068284]  do_IRQ+0x80/0xf02011/12/04 00:20:04|[379834.068286]  ret_from_intr+0x0/0xa2011/12/04 00:20:04|[379834.068288] 2011/12/04 00:20:04|[379834.068291] DMA per-cpu:2011/12/04 00:20:04|[379834.068292] CPU 0: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068297] CPU 1: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068297] CPU 2: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068297] CPU 3: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068298] CPU 4: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068299] CPU 5: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068301] CPU 6: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068302] CPU 7: hi: 0, btch: 1 usd: 02011/12/04 00:20:04|[379834.068303] DMA32 per-cpu:2011/12/04 00:20:04|[379834.068305] CPU 0: hi: 186, btch: 31 usd: 472011/12/04 00:20:04|[379834.068306] CPU 1: hi: 186, btch: 31 usd: 1402011/12/04 00:20:04|[379834.068308] CPU 2: hi: 186, btch: 31 usd: 852011/12/04 00:20:04|[379834.068309] CPU 3: hi: 186, btch: 31 usd: 1562011/12/04 00:20:04|[379834.068311] CPU 4: hi: 186, btch: 31 usd: 1002011/12/04 00:20:04|[379834.068312] CPU 5: hi: 186, btch: 31 usd: 1622011/12/04 00:20:04|[379834.068313] ...
the picture is a bit urgly, therefor:
the s200 box is direct attched to the vSphere boxes with CX4s. The vSphere boxes are attached to the switch with CX4s.
Now we upgrade the s200 with a X500 card and this card is attached to the switch with 2x SFP+.
The disconnection might be cable related. The maximum cable length for passive cables for the Intel adapters is 7 meters. (See http://www.intel.com/support/network/adapter/pro100/sb/CS-030612.htm http://www.intel.com/support/network/adapter/pro100/sb/CS-030612.htm.)
accordings to the specifications, the cables are at least SFF-8431 complient. I can not find any active/passive specification, but usualy all cables longer than 7m are active, at least they should be at a vendor like HP. Here are the specifications and a display from the switch (A1&A2). At least the GBIC in the cable is so clever to send a ID (SFP+DA1) to the switch, is this also so at passive cabes? Currently the link is up again since a couple of hours.
Thanks a lot in advance
A3 10GbE-CX4 | 2000 128 Forwarding | 0026f1-4d2480 2 Yes Yes
A4 10GbE-CX4 | 2000 128 Forwarding | 0026f1-4d2480 2 Yes Yes
Yes, passive cables will also identify themselves to the switch (and to the Intel Ethernet adpapter) as direct attached cables, so I do not think you can tell from the information that gets displayed if the cable is active or passive.
I opened at ticket at the HP professional services, but up to now the guy was unable to tell me if this is a active or passive cable,
but it's only SFP+ cable I can use, the switch supports only HP cables. So according the HP documentation it's a supported
configuration and the Intel card does not compain that the GBIC is not supported at in multiple other posts.
How to solve it? I have the feeling, that the HP guy (sitting down in east Europe) is unable to help me.
Is Intel able to help me?