Yeah, no matter how it worked before...if the Fluke tells you there is a problem, there is a problem.
So the question should not be why Aruba Transceivers causes the link to stay down (Sensitivity and Hardware could be different but FO Standards are FO Standards) but instead how exactly Cisco Transceivers have worked before (since it was reported they "worked") on that specific fiber optic run?
Original Message:
Sent: Jun 08, 2021 04:49 PM
From: Christopher Wickline
Subject: CX 6300 SFP+ Core Not Linking Fiber To Closet, Cisco 4500x Core DOES Link to Same Closet
Here's a link to the transceiver guide: It lists maximum distance and a support matrix
https://support.hpe.com/hpesc/public/docDisplay?docId=a00028947en_us
show interface transceiver threshold-violations
show interface transceiver detail
show interface x/x/x dom detail
May give you additional insight into why those issues are happening. The dom command specifically should give some more information. It should list current level, and "warning" and "alarm" levels.
As a side note, We use Fluke for our fiber testing, and we've had very good experiences. When it tells us there is a problem, there is a problem.
------------------------------
Christopher Wickline
Original Message:
Sent: Jun 08, 2021 10:50 AM
From: Alex Wright
Subject: CX 6300 SFP+ Core Not Linking Fiber To Closet, Cisco 4500x Core DOES Link to Same Closet
Hello!
We are replacing a Cisco 4500x with an Aruba 6300 SFP+ switch. When we make the swap, one closet does not come back up, 14 others do. We have tried both Cisco and Aruba access switches on the access end and neither work with the 6300, but both do work when connected to the 4500x.
Customer does not know anything about their fiber, but I can tell it is extremely old MM. Even stranger, the colors do not match for this particular run. One side is orange blue, the other is green brown, so there must be a junction box somewhere. Both sides use 1g SX(MM) SFPs, and each one is properly branded for the corresponding switch. All transceivers and fiber zips have been tested and proved to work on other connections. Both configurations are just straight trunk ports, nothing fancy. We do not even get a link light when using the aruba core. We also put a fluke tester on it and it shows the fiber to be bad, like doesn't even show a link. The distance between the buildings is well within the max range (at least if its a relatively straight shot).
Basically I think (99% sure) the fiber is bad in some way. But what I am trying to be able to explain to the customer is why it works with their existing cisco core, but not when we put in the aruba 6300. All I can think of is that there is a sensitivity threshold for allowing the link to form, and the Cisco device allows for connections to form with a weaker connection than the 6300 does, but I cannot find any data to back that up. I am also going to try using a Cisco transceiver on the aruba core with the unsupported module command in place to test this theory. Has anyone run into anything similar or have any thoughts?
Both SFPs are 1g-SFP-SX, for reference.
------------------------------
Alex Wright
------------------------------