8

All,

Recently bought 3 Cisco switches: 3560x and two 2960Xs. 3560 acting as the core switch and other 2 switches are connected using SFP+ (One SFP-H10GB-CU1M and other SFP-10G-SR)

I have issue with the SFP-10G-SR.

I connected the SFP-10G-SRs with an 50m OM3 cable. It worked initially, however the log always have:

026123: Jun 12 20:42:02.573: %SFF8472-5-THRESHOLD_VIOLATION: Te1/2: Rx power high warning; Operating value:  -0.9 dBm, Threshold value:  -1.0 dBm.
026124: Jun 12 20:48:35.730: %SFF8472-5-THRESHOLD_VIOLATION: Te1/2: Rx power high warning; Operating value:  -0.9 dBm, Threshold value:  -1.0 dBm.

In one incident, I noticed the SFP port SFP-10G-SR went down. I was playing on that so I didn't care and I simply rebooted the switches and it all came back.

Today, the SFP ports will not come up anymore. I'm in production, so I tried to change the Fibre and I tried to change SFP+ modules. No luck

show inventory

NAME: "TenGigabitEthernet1/2", DESCR: "SFP-10GBase-SR"
PID: SFP-10G-SR          , VID: V03  , SN: FNS16481363

show int te1/2

    TenGigabitEthernet1/2 is down, line protocol is down (notconnect) 
  Hardware is Ten Gigabit Ethernet, address is 24e9.b3e3.a29e (bia 24e9.b3e3.a29e)
  Description: NGPSW03 Trunk
  MTU 1500 bytes, BW 10000000 Kbit/sec, DLY 10 usec, 
     reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive not set
  Full-duplex, 10Gb/s, link type is auto, media type is SFP-10GBase-SR
  input flow-control is off, output flow-control is unsupported 
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 6d09h, output 6d09h, output hang never
  Last clearing of "show interface" counters never
  Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue: 0/40 (size/max)
  5 minute input rate 0 bits/sec, 0 packets/sec
  5 minute output rate 0 bits/sec, 0 packets/sec
     1119868 packets input, 124583435 bytes, 0 no buffer
     Received 302202 broadcasts (301488 multicasts)
     0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 301488 multicast, 0 pause input
     0 input packets with dribble condition detected
     48617012 packets output, 4642677414 bytes, 0 underruns
     0 output errors, 0 collisions, 3 interface resets
     1 unknown protocol drops
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier, 0 pause output
     0 output buffer failures, 0 output buffers swapped out

Model/Version:

Switch Ports Model              SW Version            SW Image                 
------ ----- -----              ----------            ----------               
*    1 30    WS-C3560X-24       15.0(2)SE6            C3560E-IPBASEK9-M

Any help appreciated.

Update
I tried relocating the Cable+SFP from Te1/2 to Te1/1, now the Warning is reverse.

SFF8472-5-THRESHOLD_VIOLATION: Te1/1: Rx power low warning; Operating value: -13.8 dBm, Threshold value: -13.0 dBm.
  • Do you use the same brand and type of SFP+ at both ends? One end seems to be a Finisar, probably a Cisco original If you have a look at cisco.com/c/en/us/products/collateral/interfaces-modules/… you will see that the Cisco SR SFP+ in theory transmits at 1.2dbm so it is a bit strange that the other end receives 0.9dbm. How long is the fiber that you use? Have you tried a longer cable or a 62.5 one in order to attenuate the signal? – lacasitos Sep 4 '14 at 20:53
  • Yes same model. I was using 50 meter 50/125 OM3. I just connecting a 2 meter 50/125 OM2 cable and they are not showing the error message. I tried swapping the SFPs and still no such log. So, most probably a cable issue ? BTW, The port that went down is actually in 3560x. Even if I connect the short cable to it the port is not up. Any idea how to bring the port up ? "no shut" didn't help – Firdous Amir Sep 4 '14 at 21:28
  • 2
    High power warnings are not to be ignored as this can lead to damage on the receiving end. It is possible that your optic has been damaged, have you tried replacing it? For future reference, if you are getting high power warnings, you should attenuate the signal in some way: attenuator, mandrel, longer cable, etc. – YLearn Sep 4 '14 at 21:52
  • Thanks for tip. I had reboot the 3560 to bring up the faulty port. Just tried using the same cable+SFP which created problem in the second Te port of 3560. Now, the warning is less frequent, but interestingly it says Te1/1: Rx power low alarm. That's super crazy to me. – Firdous Amir Sep 4 '14 at 22:04
  • 2
    Either you didn't clean the transceivers and fiber properly when you reconnected things (you are cleaning, right?) or the receiving optic is damaged and it's ability to receive the signal is impaired. Based on the high power warnings you received previously, I would still try to replace the optic but you can start by disconnecting and re-cleaning everything to rule that out. – YLearn Sep 4 '14 at 22:23
2

You may have two problems.

First, the SFP is transmitting out of spec. At most, it should put out -1.2dBm per the link supplied by lacasitos above. You may be able to RMA it if a Cisco SFP is complaining that another Cisco SFP is sending too hot of a signal.

Second, because you're now getting a Rx threshold error (it should not put out less than -7.3dBm and -13.8 is quite a loss on a short link), the fiber may not be fully inserted at either the tx or rx end, or the fiber is dirty or the transmitter or receiver is dirty. (It's usually the fiber getting scratched on insertion unless you're in a dirty environment.)

If RMAing is out of the question, you could always introduce some loss to bring it under spec by using an attenuator or introducing extra patches.

  • Good info. I have done more research on this as per YLearn's suggestions and found the cure. It was actually one of the SFP created the problem. If I put this SFP Module in one side of the cable (a 50 meter 50/125 OM3), it shows the high warning. If I put the same module in the other side, it showed the low power warning. I replaced this module with a different module and there more problems now. Worth mentioning, the bad module is not showing any issues when I use a short 2 meter 50/125 OM2 cable. Thanks every for help. – Firdous Amir Sep 6 '14 at 12:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.