Transceiver not recognized: how to tell if it's the coding
Your switch says unsupported transceiver, err-disables the port or the link stays down. How to read what the switch sees, and fix it with the right coding.
You've plugged in a new optic and the switch says "unsupported transceiver", err-disables the port, or simply leaves the link down. Here's how to check what the switch sees: the log, the EEPROM, the codingCoding The vendor identification in a transceiver's EEPROM. Many switches only accept optics coded for their own brand., form factorForm factor The physical shape and electrical slot of a transceiver: SFP, SFP+, SFP28, QSFP+, QSFP28 and so on. Determines which port it fits. and speed, and DOM. Then you'll know whether you need a different coding or whether the problem is somewhere else. All with read-only commands; you change nothing on your switch.
What the switch checks
An SFP, SFP+ or SFP28 carries an EEPROM laid out per SFF-8472. QSFP+ and QSFP28 use SFF-8636, QSFP-DD and OSFP the OIF's CMIS specification. It holds, among other things:
- Identifier — the first byte says which form factor the module is.
- Compliance codes and bit rate — what the module is, such as 10GBASE-LR or 25GBASE-SR.
- Vendor name, OUI and part number — this is the coding. It's what your switch looks at when it decides whether an optic is "supported".
Cisco goes further: its optics also carry a security code and checksum in the EEPROM, which the switch verifies on insertion. If that fails, you get an error and possibly an err-disabled port.
Reading the log and the EEPROM
Start with the log, not the cable. On a Cisco Catalyst a rejected optic looks like this:
%PLATFORM_PM-6-MODULE_ERRDISABLE: The inserted SFP module with interface name Gi1/1/1 is not supported
%PM-4-ERR_DISABLE: gbic-invalid error detected on Gi1/1/1, putting Gi1/1/1 in err-disable state
gbic-invalid is the err-disable cause for an invalid SFP; show interfaces status err-disabled shows which ports are in that state and why. Huawei logs an alarm that the module isn't Huawei-certified, and Arista can also put a port with an unqualified optic into errdisabled. Nothing in the log, nothing odd in the inventory, and the port just stays down? Then coding is less likely.
Next, check what the switch reads from the module:
- Cisco IOS XE —
show inventorylists the PID and serial number per optic;show interfaces Gi1/1/1 transceivershows the DOM values the switch reads from the module. - Juniper Junos —
show chassis hardwarelists optics from another vendor asNON-JNPR, sometimes withUNSUPPORTED.show chassis pic fpc-slot 0 pic-slot 0gives the Xcvr vendor and vendor part number. - Arista EOS —
show inventorylists manufacturer, model and serial number per transceiver slot. - Huawei VRP —
display transceiver interface GigabitEthernet0/0/1 verboseincludes the manufacturer details. - MikroTik RouterOS —
/interface ethernet monitor sfp-sfpplus1showssfp-vendor-name,sfp-vendor-part-number,sfp-typeandeeprom-checksum. If that reads bad, you won't get DOM either. MikroTik documents no vendor restriction, so look at speed or the checksum first.
Matching the coding to your platform
Compare that with what your platform supports. Cisco has show interfaces transceiver supported-list on the switch and the Transceiver Module Group matrix online; Juniper has the Hardware Compatibility Tool. Watch two things:
- Brand — an optic coded for Juniper isn't recognized as a Cisco optic by a Cisco switch, even if the hardware is identical.
- Platform and software release — even within one brand, not every optic is supported on every platform. Cisco has even had bugs where a genuine optic was flagged as gbic-invalid.
There are commands that turn the check off; which ones per platform is in enable third-party optics. They disable the check, not the consequences: support may ask you for an original or qualified module, and after an upgrade or RMA it can start all over. The right coding is the real fix.
Not a coding problem? Form factor, speed and DOM
Not every failure is coding. Three classics:
- 10G optic in an SFP28 port — fits and often works, but not on every platform. Sometimes you set the speed to 10G yourself, and sometimes every port in a group must run the same speed. One 10G optic among three 25G ports is then asking for a port that stays down.
- 25G optic in an SFP+ port — it fits, but an SFP+ port tops out at 10G. That only works if the optic's datasheet lists 10G as well; otherwise use a 10G optic.
- Breakout — a QSFP28 with a 4x25G breakoutBreakout One high-speed port split into several slower ones, for example a 100G QSFP28 into 4× 25G SFP28 with a breakout cable. The switch must support it. doesn't split itself. The breakout has to be in the switch configuration, and how you do that differs per platform.
If the switch shows DOM values, it can at least read the module. No unsupported message, no err-disable, and the link still stays down or starts flapping? Then the problem is most likely in the optical path (Rx power, reach, a dirty connector) or in settings such as speed and FEC. See reading DOM/DDM.
Checklist: transceiver not recognized
- Read the log and note the exact message and the err-disable cause.
- Read vendor, part number and type from the EEPROM with a read-only command.
- Compare the coding with the compatibility matrix for your platform and software release.
- Check that the form factor and speed of optic and port match.
- Check, for a breakout, that it's in the switch configuration.
- Read DOM and look at Tx and Rx power if the module has been accepted.
- Order an optic with your platform's coding if the coding doesn't match.
What to do now
If the coding doesn't match, the fix is simple: an optic with the right coding. With our optics, such as the SFP-10G and our 25G SFP28 modules, you choose the coding when you order. They come from Solid Optics, a Dutch company that tests them in the Netherlands, on customer equipment too. Every module carries a lifetime warranty. In Amsterdam we deliver same day. Unsure about the coding, or want a module tested first? See our services or send us your switch model and software release.

Founder of FiberNxt and Xyphen IT
Yvan Watchman is the founder of FiberNxt and Xyphen IT. He has been building and running networks and data centre infrastructure in and around Amsterdam for years.
Related products and services
Frequently asked questions
What does gbic-invalid mean on a Cisco switch?
The switch has read the transceiver and doesn't accept it as a valid or supported module, so it err-disables the port. Usually it's the coding in the EEPROM; occasionally it's a software bug, even with a genuine optic. Start by checking what the switch reads from the module.
Can a transceiver fit the port and still not work?
Yes. An SFP+ fits an SFP28 port, for example, but coding, speed and port configuration have to match too. On some switches you have to set an SFP28 port to 10G yourself first.
Can you supply optics with my switch's coding?
Yes. You pick the coding when you order, from generic MSA to Cisco, Juniper, Arista, Huawei, MikroTik and about fifteen other brands. Not sure which one you need? Send us the switch model and software version.