[c-nsp] Catalyst 3750-X with SFP-10G-ZR transceiver - troubleshooting

[Joe Crap]

We need to run 10GbE dark fiber between two 3750-X switches (each with a C3KX Service Module installed) over a 30km distance where the measured insertion loss makes ER transceivers marginal (total loss > 12dB) - so we're forced to use ZR transceivers with their greater reach. As of the Aug 28th release of IOS 15.2(1)E the ZR optics are now supported in this switch - however our initial testing using 3rd-party transceivers has been unsuccessful: we are unable to get a link light on the port and the logical port never comes up when the fiber is plugged in. I was hoping someone could help me sanity check the testing methodology?
We're using 3rd-party SFP-10G-ZR clones that report VID: V01 to the switch. They support DOM, and we've also tested LR transceivers from the same 3rd-party successfully. My testing has been pretty simple so far:
Connect Tx-to-Rx and vice versa between the two transceivers using a 2m duplex OS1 SMF patch lead that has 15dB fixed inline attenuators connected to bring the received signal strength down to the supported range (-6 to -27 dBm)
Connected this way however we are unable to get a link light or port-up state on these ports. The ZR transceivers show up in a "show inv" output, and the DOM on the transceivers is reporting a valid strength Rx signal:
switch1#sho int te1/1/2 trans detailITU Channel not available (Wavelength not available),Transceiver is internally calibrated.mA: milliamperes, dBm: decibels (milliwatts), NA or N/A: not applicable.++ : high alarm, +  : high warning, -  : low warning, -- : low alarm.A2D readouts (if they differ), are reported in parentheses.The threshold values are calibrated.                               High Alarm  High Warn  Low Warn   Low Alarm          Temperature         Threshold   Threshold  Threshold  ThresholdPort       (Celsius)          (Celsius)   (Celsius)  (Celsius)  (Celsius)--------- ------------------  ----------  ---------  ---------  ---------Te1/1/2     44.4                75.0        70.0         0.0       -5.0                               High Alarm  High Warn  Low Warn   Low Alarm           Voltage            Threshold   Threshold  Threshold  ThresholdPort       (Volts)            (Volts)     (Volts)    (Volts)    (Volts)---------  ---------------    ----------  ---------  ---------  ---------Te1/1/2    3.27                  3.63        3.46        3.13       2.97            Optical            High Alarm  High Warn  Low Warn   Low Alarm           Transmit Power     Threshold   Threshold  Threshold  ThresholdPort       (dBm)              (dBm)       (dBm)      (dBm)      (dBm)---------  -----------------  ----------  ---------  ---------  ---------Te1/1/2      1.7                 4.7         4.0         0.0       -0.9            Optical            High Alarm  High Warn  Low Warn   Low Alarm           Receive Power      Threshold   Threshold  Threshold  ThresholdPort       (dBm)              (dBm)       (dBm)      (dBm)      (dBm)-------    -----------------  ----------  ---------  ---------  ---------Te1/1/2    -12.2                -5.9        -7.2       -25.4      -27.2
If we take the same patch cable, remove the attenuators and plug it into the other LR transceivers in the same switches they get a link light instantly. I realise these two sets of transceivers are operating at different frequencies (1310nm vs 1550nm) but afaik an OS1 cable should be fine with both, right? We've also swapped the ZR transceivers into the same ports the LRs are working fine in, it makes no difference.
I'm afraid my own fiber knowledge is quite limited (much of it learned from trying to troubleshoot this!), but can anyone suggest further troubleshooting steps? We don't have the equipment to run an OTDR test on the patch leads, and I'm trusting the DOM reporting to be at least vaguely accurate for signal strength.
Thanks in advance for any assistance!
- J