Fryguy's Blog Back to Tools
Paste the output of show interfaces diagnostics optics (Junos) or show controllers <if> [all|phy] (Cisco IOS-XR) and get a readable report — Tx/Rx light levels, laser bias, module temperature and voltage, each plotted against the alarm and warning thresholds your own optic reports, with the alarm flags and plain-English findings pulled out. Handles single-lane SFP/SFP+, multi-lane QSFP (Lane N), and output for many interfaces at once.
⚠ Supports Junos and Cisco IOS-XR (
show controllers <if> [all|phy]) today — format is auto-detected, just paste either one. I'm working on adding IOS-XE, NX-OS, Arista, and other platforms as I get valid sample output to build and test against.Your optics report will appear here…
Judgments come from the thresholds and alarm flags the module itself reports, not from a generic rule of thumb — every optic type has different limits. If you type the far end's Tx power, the report also shows the implied link loss (far-end Tx minus local Rx, in dB). Built against typical Junos output (EX/QFX/MX/PTX) and Cisco IOS-XR
show controllers output (ASR9k).Examples (click to load)
Tx powerLight this optic is sending (
Laser output power). Low or missing Tx points at the local optic or its own connector.Rx powerLight arriving from the far end (
Receiver signal average optical power / Laser rx power). The number that tells you about the fiber path and the far end.BiasCurrent driving the laser. It creeps up as the laser ages to hold output steady, so a climbing bias plus falling Tx means end of life.
Warning vs alarmWarning thresholds sit inside the alarm thresholds. A warning means “look soon”; an alarm means out of spec.
dBm vs mWSame measurement two ways. dBm = 10×log10(mW), so 0 dBm = 1 mW and every 3 dB is a doubling or halving.
- Inf dBmZero measured light (0 mW). Treated as “no light”.
Approximate IEEE 802.3 average-power figures, per lane where noted. Real optics vary by vendor — the thresholds your module reports (which this tool uses) are the authority; use this only for a sanity check.
| Optic | Tx power | Rx range |
|---|---|---|
| 1000BASE-SX | -9.5 to -4 dBm | -17 to 0 dBm |
| 1000BASE-LX | -9.5 to -3 dBm | -19 to -3 dBm |
| 10GBASE-SR | -7.3 to -1 dBm | about -9.9 to -1 dBm |
| 10GBASE-LR | -8.2 to +0.5 dBm | -14.4 to +0.5 dBm |
| 10GBASE-ER | -4.7 to +4 dBm | -15.8 to -1 dBm |
| 10GBASE-ZR | 0 to +4 dBm | -24 to -7 dBm |
| 100GBASE-SR4 (per lane) | -8.4 to +2.4 dBm | -10.3 to +2.4 dBm |
| 100GBASE-LR4 (per lane) | -4.3 to +4.5 dBm | about -10.6 to +4.5 dBm |
No Rx lightFar-end laser off or port admin-down, unplugged/cut fiber, Tx and Rx swapped, wrong fiber type or wavelength, or a filthy connector. Loop the optic back to prove it.
Rx low, Tx fineThis optic is healthy; loss is on the path or the far end. Clean both ends, check patch cords, bends, splices and any patch panel in between.
Rx too highShort run with a long-reach optic can overload the receiver. Add an inline attenuator (commonly 5–10 dB) at the receiving end.
Tx lowDirty local connector or an aging laser. Clean and re-read; if it stays low, replace the optic.
Bias highLaser is being driven hard to keep power up — approaching end of life. Plan a swap before it fails.
One weak laneOn QSFP/MPO links, a single lane far below its siblings usually means one strand or one MPO position is dirty or damaged.
Light OK, link downNot an optics problem: check speed/FEC settings, optic type mismatch, admin state, or a genuine far-end fault. Rx losses of a fraction of a dB between reads are normal.
Provided as-is, with no warranty of any kind. This runs entirely in your browser — nothing you paste is uploaded anywhere. It is a convenience tool, not a certified reference; always double-check before acting on the results, especially on production links. See the full disclaimer.