Probably there isn't (on AOS-S), I mean...a show interface brief should show you the physical ethernet interfaces belonging to each specific Trk<id> interface, if any...and there you will see the member speeds, then you need to perform the arithmetic sum to figure out the aggregated speed of a particular Trk<id>. On AOS-CX things are a little bit different.
Original Message:
Sent: Nov 29, 2024 12:38 PM
From: t.antony
Subject: FCS errors on fiber transceiver
Yes, I tried 2 commands sh interfaces trk1 & show interfaces ethernet trk1 but they both return the same output.
I also checked on the Web UI, but it doesn't show any lacp, only VSF stacking / logging.
sh interfaces trk1
Status and Counters - Port Counters for trunk Trk1 Name : Trk1 MAC Address : a40e75-679dc0 Link Status : Up Port Enabled : Yes Totals (Since boot or last clear) : Bytes Rx : 2,406,643,545 Bytes Tx : 4,449,615,990 Unicast Rx : 52,955,491 Unicast Tx : 54,937,528 Bcast/Mcast Rx : 18,491,634 Bcast/Mcast Tx : 11,968,722 Errors (Since boot or last clear) : FCS Rx : 2 Drops Tx : 0 Alignment Rx : 0 Collisions Tx : 0 Runts Rx : 0 Late Colln Tx : 0 Giants Rx : 0 Excessive Colln : 0 Total Rx Errors : 2 Deferred Tx : 0 Others (Since boot or last clear) : Discard Rx : 0 Out Queue Len : 0 Unknown Protos : 0 Rates (5 minute weighted average) : Total Rx (bps) : 1,360,000 Total Tx (bps) : 816,000 Unicast Rx (Pkts/sec) : 279 Unicast Tx (Pkts/sec) : 250 B/Mcast Rx (Pkts/sec) : 18 B/Mcast Tx (Pkts/sec) : 5 Utilization Rx : 00.06 % Utilization Tx : 00.03 %
show interfaces ethernet trk1
Status and Counters - Port Counters for trunk Trk1 Name : Trk1 MAC Address : a40e75-679dc0 Link Status : Up Port Enabled : Yes Totals (Since boot or last clear) : Bytes Rx : 2,437,395,759 Bytes Tx : 4,460,321,673 Unicast Rx : 52,996,134 Unicast Tx : 54,968,961 Bcast/Mcast Rx : 18,493,773 Bcast/Mcast Tx : 11,969,211 Errors (Since boot or last clear) : FCS Rx : 2 Drops Tx : 0 Alignment Rx : 0 Collisions Tx : 0 Runts Rx : 0 Late Colln Tx : 0 Giants Rx : 0 Excessive Colln : 0 Total Rx Errors : 2 Deferred Tx : 0 Others (Since boot or last clear) : Discard Rx : 0 Out Queue Len : 0 Unknown Protos : 0 Rates (5 minute weighted average) : Total Rx (bps) : 1,584,000 Total Tx (bps) : 792,000 Unicast Rx (Pkts/sec) : 282 Unicast Tx (Pkts/sec) : 248 B/Mcast Rx (Pkts/sec) : 19 B/Mcast Tx (Pkts/sec) : 3 Utilization Rx : 00.07 % Utilization Tx : 00.03 %
Original Message:
Sent: Nov 29, 2024 12:11 PM
From: parnassus
Subject: FCS errors on fiber transceiver
Hi, glad you found a solution that is working for your scenario.
"How can I see the speed of trk1"
Have you tried with a simple show interface ethernet Trk1 ? IIRC it should show you the aggregate (logical) interface's bandwidth.
Original Message:
Sent: 11/29/2024 8:12:00 AM
From: t.antony
Subject: RE: FCS errors on fiber transceiver
I setup the trunk, and now I'm showing this, and everything works. No packet loss, and when I unplug one cable, the connection stayed up.
Its using the default L3 level.
Core-SW-1# sh trunks 27-28 Load Balancing Method: L3-based (default) Port | Name Type | Group Type ------ + -------------------------------- ---------- + ----- -------- 27 | New Bldg Tunk LAG Trk1 1000LX | Trk1 LACP 28 | New Bldg Tunk LAG Trk1 1000LX | Trk1 LACP
How can I see the speed of trk1? When I do sh lacp counters I get this below.
LACP Port Counters. LACP LACP Marker Marker Marker Marker Port Trunk PDUs Tx PDUs Rx Req. Tx Req. Rx Resp. Tx Resp. Rx Error ---- ------ --------- --------- -------- -------- -------- -------- -------- 27 Trk1 1512 1512 0 0 0 0 0 28 Trk1 1505 1503 0 0 0 0 0
Original Message:
Sent: Nov 25, 2024 06:55 PM
From: t.antony
Subject: FCS errors on fiber transceiver
Thank you
Original Message:
Sent: Nov 25, 2024 11:46 AM
From: parnassus
Subject: FCS errors on fiber transceiver
Hi, yes...if you aggregate (Port Trunking = LAG with or without LACP as control protocol, with LACP is often referred to Dynamic while without LACP is often referred to Static or Non-Protocol) two or more physical interfaces the resulting aggregated bandwidth will be the sum (say 4 links x 1 Gbps/link -> aggregated bandwidth: 4 Gbps for the entire logical interface) but this doesn't translate into a 4 Gbps egress bandwitdh for any communication stream (each outgoing stream at best will use one LAG physical channel at time so one physical port at time and, at the end, it outgoing speed will be limited by the port's bandwidth).
Yes, the examples you did are good (SRC 49+50 2 x 1 Gbps <==> DST 25+26 2 x 1 Gbps)
"So if one of the ports go down in lcap, it will only degrade it to 1 Gb correct? But the interface will stay up?" both are correct.
Original Message:
Sent: Nov 25, 2024 08:34 AM
From: t.antony
Subject: FCS errors on fiber transceiver
Is this how I make lcap on both switches? Can I call it from trk1 to something meaning like 'building lcap'? All my switch ports are 1 Gb ports. So this means the trunk speed will be about 2 Gb? This will also do load balancing or is it additional I need to setup after creating the LAG? So if one of the ports go down in lcap, it will only degrade it to 1 Gb correct? But the interface will stay up?
Origin switch:
trunk 49,50 trk1 lacp
interface trk1
tagged vlan 1,2-5
Destination switch:
trunk 25,26 trk1 lacp
interface trk1
tagged vlan 1,2-5
Original Message:
Sent: Nov 22, 2024 04:45 PM
From: Davide Poletto
Subject: FCS errors on fiber transceiver
Hi! links aggregation (a LAG) requires - to be implemented (both at origin and at destination) - to deal with a single logical entity (which normally is represented by a "single switch" or by a "virtual switch" or by a "cluster made of a pair of switches", in this latter case the cluster - like the VSX clustering technology available on ArubaOS-CX based switched - should support the concept of multi-chassis LAG or MC-LAG).
You can't form a LAG logical interface with member physical interfaces belonging to different standalone switches (say member of pair: Switch A1 and A2 or pair: Switch B1 and B2).
The Aruba 2930F supports VSF...implementing it could be a way to overcome the above LAG-related restriction (but require some work/materials and VSF setup)...or you can play with your current setup and work with RSTP planning to let the STP to keep the A2 to B2 link down (so A2 - A1 ---- B1 - B2 will form a chain), a link that would be reactivated by STP if the A1 to B1 link goes down.
Original Message:
Sent: 11/22/2024 3:58:00 PM
From: t.antony
Subject: RE: FCS errors on fiber transceiver
Thank you, after drawing a network diagram it does seem with my method, I'm creating a loop.
With the Port Trunk idea, can I do Link Aggregation between 2 interfaces on different switches? Wouldn't that also possibly cause a loop between the 2 switches? For example Link Aggregation between 10.0.0.6 interface 20 & 10.0.0.4 interface 21? If yes, then if 10.0.0.6 interface 20 goes down for some reason, the connection will be kept alive by 10.0.0.4 interface 21 correct?
I have done Link Aggregation before, but with ports on the same switch.
Original Message:
Sent: Nov 22, 2024 03:51 PM
From: parnassus
Subject: FCS errors on fiber transceiver
Hi! if you "like to add a second backup connection between the 2 buildings" the most simple thing to do is to create a Port Trunk (which, in ArubaOS jargon, means forming a logical interface representing a Links Aggregation = physical interfaces with identical characteristics aggregated together to form a logical interface providing links redundancy and egress traffic load balancing), a Port Trunk - named Trk<id> - on each switch and so use - aggregated - both yours 1 Gbps fiber optic long range links.
The other approach (your one?) seems to be prone to create a network loop between the two Buildings A and B and not simply an active link with a backup (disabled) link (since Switch A1 and Switch B1, already interconnected with a single link, become part of a loop which closes when you interconnect Switch A2 and Switch B2...provided that you - reasonably - have A1 already connected to A2 on Building A and B1 already connected to B2 on Buildind B <- here I'm implying that you are working with standalone switches and not Stacked ones) and this network loop should be broken by Spanning Tree which will immediately kicking in to keep your network topologically loop-free.
Original Message:
Sent: 11/22/2024 3:19:00 PM
From: t.antony
Subject: FCS errors on fiber transceiver
I'm connecting our 2 buildings using fiber transceiver. My existing connection is between 2 JL261A switches (10.0.0.6 & 10.0.0.9) using transceiver model J4859D / part number 1990-3656. Same on both switches.
I like to add a second backup connection between the 2 buildings. The 2 switches I want to use are also JL261A switches (10.0.0.4 & 10.0.0.1) using transceiver model J4859D / part number 1990-3656. Same on both switches. But when I connected the yellow fiber optic cable between the 2 switches, I was not getting a link light. So on one of the switch, I had to flip the fiber optic cable. I got a link light that time. My computer is connected to the 10.0.0.1 switch on port 1. I noticed my computer lost connection to network, but the phone stayed up. To test I even made a phone call and it went through. The VoIP phone is connected to switch port, and the computer is connected to phone. What could be the cause of this?
When I looked on 10.0.0.1 port 27, I saw some FCS errors.
So all 4 switches are the same model and the same transceivers. All 4 tagged with same VLANs. This is the transceiver I'm using on all 4 switches.
https://www.fs.com/products/92107.html