مختلف وینڈرز کے ہاں پورٹس کے نام اور LED اشارے
ہماری ٹیم ترجمے پر کام کر رہی ہے۔ یہ مضمون عارضی طور پر اردو میں دستیاب نہیں ہے، اس لیے اسے انگریزی میں دکھایا جا رہا ہے۔
"Put the module in Et33/1" means nothing on a Juniper and "xe-0/0/48" nothing on a Nexus; the same physical cage has a different name in every network OS, and breakout children, stack members and line cards add their own notation. The LEDs next to the cage are a second language: green, amber, blinking, off — with vendor-specific meanings. This page gives the naming grammar per NOS, the slot/port semantics behind it, the breakout and management-interface conventions, the LED codes, and the commands that make a port blink so you can find it in a rack.
Naming grammar by NOS
| NOS | Pattern | Examples | Read as |
|---|---|---|---|
| Cisco IOS / IOS-XE | <Type><switch>/<module>/<port> | GigabitEthernet1/0/1, TenGigabitEthernet1/1/1, TwentyFiveGigE1/0/25, FortyGigabitEthernet1/1/1, HundredGigE1/0/49, Gi1/0/1 (short) | switch (stack member) / module (0 = built-in, 1 = uplink module) / port |
| Cisco NX-OS | Ethernet<slot>/<port>[/<sub>] | Ethernet1/1, Eth1/49/1 (breakout child) | slot (1 on fixed) / port / breakout lane |
| Cisco IOS-XR | <Type><rack>/<slot>/<instance>/<port>[/<sub>] | HundredGigE0/0/0/0, FourHundredGigE0/0/0/12, TenGigE0/0/0/0/1 (breakout) | rack / slot / instance (subslot) / port |
| Arista EOS | Ethernet<port>[/<lane>] | Ethernet1, Ethernet33/1…33/4 (breakout children), Et49/1 | port / lane; a QSFP port without breakout is Et33/1 |
| Juniper Junos | <media>-<fpc>/<pic>/<port>[:<channel>] | ge-0/0/0, xe-0/0/48, et-0/0/50, et-0/0/50:1 (breakout), mge- (multigig), fte- (40G on EX4300) | media (ge 1G, xe 10G, et 25G+) - FPC (stack member/line card) / PIC (module) / port |
| Huawei VRP | <Type><chassis>/<slot>/<port> | GigabitEthernet0/0/1, XGigabitEthernet0/0/1, 25GE1/0/1, 100GE1/0/1, 100GE1/0/1:1 (breakout) | chassis / slot / port |
| H3C Comware | <Type><chassis>/<slot>/<port> | GigabitEthernet1/0/1, Ten-GigabitEthernet1/0/49, Twenty-FiveGigE1/0/1, HundredGigE1/0/53:1 | as VRP |
| HPE Aruba AOS-CX | <member>/<slot>/<port> | 1/1/1, 1/1/49:1 (breakout) | member / slot / port |
| HPE Aruba AOS-S (ProCurve) | <port> or <letter><port> | 1, A1, 49 | flat; letters for modules |
| Dell OS10 | ethernet<node>/<slot>/<port>[:<sub>] | ethernet1/1/1, ethernet1/1/49:1 | node / slot / port : breakout |
| Extreme EXOS | <slot>:<port> | 1:1, 2:49 | slot (stack member) : port |
| Ruckus ICX | <unit>/<slot>/<port> | 1/1/1, 1/2/1 (uplink module) | unit / slot / port |
| Brocade FOS | <slot>/<port> or port index | 0, 1/3 | index 0-based on fixed switches |
| SONiC | Ethernet<lane-index> | Ethernet0, Ethernet4, Ethernet8… (4-lane platform), Ethernet0…Ethernet3 after 4× breakout; 8-lane: Ethernet0, Ethernet8, Ethernet16 | first lane index, not port number; aliases like Eth1/1 or etp1 exist per platform |
| Cumulus / NVUE | swp<port>[s<lane>] | swp1, swp32, swp32s0…swp32s3 (breakout), eth0 (mgmt) | switch port / sub-port |
| MikroTik RouterOS | <type>-<n> | ether1, sfp-sfpplus1, sfp28-1, qsfp28-1-1…-1-4 (lanes) | type - number - lane |
| Ubiquiti | <port> in UI, eth<n> inside | Port 49, eth48 | 0-based inside |
| Nvidia Onyx (MLNX-OS) | Eth<slot>/<port>[/<sub>] | Eth1/1, Eth1/1/1 | — |
| Linux hosts | eth<n>, ens<n>f<n>, enp<bus>s<slot>f<fn> | enp94s0f1, ens3f0np0 | predictable names from PCIe address (NIC tools) |
Slot, module, port: what the numbers mean
| Number | On a fixed switch | On a stack | On a chassis |
|---|---|---|---|
| First (switch / chassis / FPC / member / node) | 1 (or 0 on Juniper/Huawei) | stack member ID 1–8 | line-card slot (Juniper FPC) or chassis number |
| Second (module / slot / PIC) | 0 = built-in ports, 1 = uplink module (Cisco), or 0 always | same | subslot/PIC on a card |
| Third (port) | front-panel number, 1-based (0-based on Juniper, SONiC, FOS) | — | — |
| Fourth (sub / lane / channel) | breakout child 1–4 or 0–3, or 1–8 | — | — |
Port numbering on the faceplate is usually left-to-right, top row odd and bottom row even (1/2, 3/4…) on dual-row copper blocks, and left-to-right sequential on SFP rows — check the silkscreen, not the assumption.
Breakout children
| NOS | Parent | Children | Enabling command |
|---|---|---|---|
| NX-OS | Ethernet1/49 | Ethernet1/49/1–/4 | interface breakout module 1 port 49 map 25g-4x |
| Arista | Ethernet49/1 | Ethernet49/1–49/4 | speed on children / hardware port-group |
| Junos | et-0/0/48 | et-0/0/48:0–:3 | set chassis fpc 0 pic 0 port 48 channel-speed 25g |
| IOS-XR | HundredGigE0/0/0/0 | TwentyFiveGigE0/0/0/0/0–/3 | controller Optics … breakout 4x25 |
| Huawei | 100GE1/0/1 | 25GE1/0/1:1–:4 | port split dimension … |
| Dell OS10 | ethernet1/1/1 | ethernet1/1/1:1–:4 | interface breakout 1/1/1 map 25g-4x |
| SONiC | Ethernet0 | Ethernet0, Ethernet1, Ethernet2, Ethernet3 | config interface breakout Ethernet0 4x25G |
| Cumulus | swp1 | swp1s0–s3 | nv set interface swp1 link breakout 4x25G |
| MikroTik | qsfp28-1-1 | qsfp28-1-1…-1-4 always present | speed per lane |
Commands in context: Port configuration recipes; cabling: Breakout & MPO.
Management and special interfaces
| Interface | Names |
|---|---|
| Out-of-band Ethernet | mgmt0 (NX-OS), GigabitEthernet0/0 (IOS-XE), MgmtEth0/RP0/CPU0/0 (XR), Management1 (EOS), me0/em0/fxp0/vme (Junos), MEth0/0/1 (Huawei), M-GigabitEthernet0/0/0 (H3C), mgmt (AOS-CX), eth0 (SONiC, Cumulus), ether1 often (MikroTik) |
| Console | con0, console |
| Loopback / SVI / VLAN | Loopback0, Vlan10, irb.10 (Junos), Vlanif10 (Huawei) |
| LAG | Port-channel1 (Cisco), ae0 (Junos), Eth-Trunk1 (Huawei), Bridge-Aggregation1 (H3C), lag 1 (AOS-CX), port-channel1 (OS10), PortChannel0001 (SONiC), bond0 (Cumulus) (Forwarding basics) |
| Stack / VC ports | vcp-255/0/50 (Junos), StackPort1/1 (Cisco), stack-port 1/1 (Stacking & MLAG) |
LEDs
| LED | State | Usual meaning | Vendor variations |
|---|---|---|---|
| Port link | off | no link, port shut, or module not accepted | — |
| Port link | solid green | link up at top speed | Cisco: green = link; Arista: green |
| Port link | blinking green | link up with traffic | most |
| Port link | solid amber / orange | link at lower speed (1G in a 10G port), or errdisabled/blocked | Cisco IOS: amber = disabled/errdisable; Juniper: amber = duplex/half or fault; Aruba: orange = reduced speed |
| Port link | blinking amber | errdisable, PoE fault, or STP blocking | Cisco |
| Port link | alternating green/amber | link fault / diagnostics | some |
| QSFP-DD / OSFP port LEDs (2 or 4 per cage) | per-child status after breakout | one LED per breakout child on many platforms | Arista, Cisco N9K |
| Module status LED | amber | module unsupported/invalid | Nexus, some Juniper |
| Speed LED (separate, when present) | off / green / amber | 1G / 10G / 25G etc. per mode button | Cisco Catalyst "mode" button toggles port LEDs between STAT, SPEED, DUPLEX, STACK, PoE |
| System / status | green / amber / red | OK / warning (fan, temperature) / fault | — |
| Stack / master | solid / blinking | active member / stack master | Cisco, Extreme |
| PSU / fan | green / amber / off | OK / failed / absent | — |
| Beacon / locator | blue blinking | identifies the chassis | hw-module beacon, system beacon |
LED codes are in each platform's hardware installation guide; the "mode" button on Catalyst-class switches changes what the port LEDs show, which is a classic source of misread states.
Finding a port physically
| NOS | Command |
|---|---|
| Cisco NX-OS | locator-led interface ethernet 1/1 |
| Cisco IOS-XE | hw-module beacon slot 1 on (chassis), port LED by mode button |
| Arista | locator-led interface Ethernet1 on / show locator-led |
| Junos | request chassis beacon fpc 0 port 5 pic 0 on |
| Huawei | display device + physical labels; beacon on chassis |
| Dell OS10 | beacon enable ethernet 1/1/1 (platform-dependent) |
| SONiC | sfputil show presence; platform-specific LED control |
| Linux host | ethtool -p eth0 10 (blink 10 s) |
In CodingBox
The bench does not care what the port is called — but the field notes in the code database do: recording "Et49/1 on spine-2, breakout child 2" next to a module's serial number links a rejected-module event to the exact cage, which is what lets a coding or policy problem be reproduced.