I’ve a problem getting 10G fiber connection to work with this hardware:
two fiber odules of type S+85DLC03D from eurodk.com
one cable - LC to LC - OM3 - 10 meter - part# 1170749 from allekabels.nl
No matter how I switch modules and/or cables, I can’t get the connection to work: only one end has LEDs turning on (when on, both ACT/LINK and 10G are both on), the other end has the LEDs off.
So:
How can I check if this is because of the cable or a module failure?
If it is a module: how can I figure out which one?
What do the values (like “Rx Lose”) in the dialog mean?
The screenshots show a left and right image with the SFP status of the first/second SFP+ port on my CRS226-24G-2S+RM on firmware v6.36.
Situation A
Left = LEDs - SK151215M81288 - Rx Lose blank - running - link OK - Tx Bias Current blank
Right = off - SK151215M81289 - Rx Lose checked - no running - no link - Tx Bias Current 2mA
you seem to be using the right cable (125/50um multimode) with the right modules (multimode sfp+ with 300m operating range)
first of all, make sure, you run 6.37rc, otherwise readings on your CCR2xx may be wrong (rx,tx levels are not refreshed accordingly)
sit B and sit C has surprisingly same readings, which is because of this bug. if you insert/remove the module, the reading will be
refreshed. but simply disconnect/reconnect the cable will not result any change in the measured values. this problem is
addressed in the recent RC (see release notes).
“rx lose” sounds to be a typo for me. it should have been “rx loss”. that means no incoming signal is detected.
at the “sfpplus2” port i see on the capture transmit power of -31dBm. that’s nothing. either the reading is bad,
or the module is not transmitting. tx values shall be somewhere between -1dBm and -6dBm.
whereas comparable multimode SFP+ transceivers have receiver sensitivity of >-11..-15dBm, that means the reading
on your screengrab (-16dBm) is out of the valid range, so the module can’t interpret the received information.
hence the “rx loss” checkbox.
swap the modules and keep the cable as it is. this will tell you who is the bad guy: the cable or one of the modules.
if the readings stay the same, then the cable is to blame.
if the readings swap, it is one of the modules.
problem can be physical (cable/module connector is dirty) or hw issue (fiber in cable is broken, tosa/rosa in SFP is broken)
I cold booted the device (it’s not a CCR but a CRS226-24G-2S+RM on firmware v6.36) and added a screenshot of that situation in the old post. It’s very similar the same as C: only slight alterations of values except for “Rx Power” on “sfpplus2” which is now a tenth of what it was.
“rx lose” sounds to be a typo for me. it should have been “rx loss”. that means no incoming signal is detected.
at the “sfpplus2” port i see on the capture transmit power of -31dBm. that’s nothing. either the reading is bad,
or the module is not transmitting. tx values shall be somewhere between -1dBm and -6dBm.
whereas comparable multimode SFP+ transceivers have receiver sensitivity of >-11..-15dBm, that means the reading
on your screengrab (-16dBm) is out of the valid range, so the module can’t interpret the received information.
hence the “rx loss” checkbox.
swap the modules and keep the cable as it is. this will tell you who is the bad guy: the cable or one of the modules.
if the readings stay the same, then the cable is to blame.
if the readings swap, it is one of the modules.
That’s what I did.
Am I right that one of the modules is broken?
If so: how can I tell which module is broken?
And if so: if it’s a simple fixable thing like dirt?
problem can be physical (cable/module connector is dirty) or hw issue (fiber in cable is broken, tosa/rosa in SFP is broken)
Modules and cables are brand new, I carefully removed the rubber caps and inserts, then connected the cables to the SFP+ modules so I don’t think I broke anything.
