Cheap 8 * 2.5Gbit Managed Switches

[jauling]

Hope it's okay to (attempt to) revive this thread?

I just bit the bullet and pulled the trigger on two switches on AliExpress. Decent discounts and free shipping meant I got these together for €70.

Goodtop 8x 2.5G + 2x 10G SFP+ Managed ZX310S-8T2XS
Goodtop 5x 2.5G + 1x 10G SFP+ Unmanaged GT-ST015

The ZX310S-8T2XS seems to be the same model# used by others. Goodtop hasn't listed any firmware updates for it yet, they told me via chat it's MaxLinear. I'll pop it open to confirm upon receipt. This switch will be LACP'd to my existing main 1G switch. Eventual plan is to do a short run fiber upstairs to the GT-ST015. Need to find the time and patience to replace the CAT6 run with fiber, and hope the conduit bends aren't too aggressive. I'd like to avoid 10GBase-T purely due to higher power consumption.

 

[Shonk]

I Picked up a GoodTop 4x2.5G 1x10G RJ45 + 1x10G SFP+ the other day with some other stuff
Its ZX310S-4T2XH
Platform = RTL8372N + RTL8261BE (May have been replaced with RTL8261C as its 12-25 build date)
OEM = Shenzhen HongRui Optical Technology
Firmware V200.1.15
Doesnt support any diagnostics on the SFP+ Port

the newer Realtek SDK on these doesnt seem to support the hidden menu's and such

ftcolor.cgi
ftdft.cgi
ftlogo.cgi
menu_ft.cgi
or
oem_set.html
system_set.html

It seems solid enough
Using it with an LR Optical uplink to my son's room then 10G to his PC via the 10G RJ45

 

[jauling]

Hope it's okay to (attempt to) revive this thread?

I just bit the bullet and pulled the trigger on two switches on AliExpress. Decent discounts and free shipping meant I got these together for €70.

Goodtop 8x 2.5G + 2x 10G SFP+ Managed ZX310S-8T2XS
Goodtop 5x 2.5G + 1x 10G SFP+ Unmanaged GT-ST015

The ZX310S-8T2XS seems to be the same model# used by others. Goodtop hasn't listed any firmware updates for it yet, they told me via chat it's MaxLinear. I'll pop it open to confirm upon receipt. This switch will be LACP'd to my existing main 1G switch. Eventual plan is to do a short run fiber upstairs to the GT-ST015. Need to find the time and patience to replace the CAT6 run with fiber, and hope the conduit bends aren't too aggressive. I'd like to avoid 10GBase-T purely due to higher power consumption.

Received both switches yesterday. Quite acceptable shipping wait, approximately a week. The GT-ST015 is still in the box, but I've configured the ZX310S-8T2XS and it's wall mounted live in my LAN now. Running two interface LACP aggregation to my existing 1G switch, and moved a couple devices over to 2.5G now. The web interface is surprisingly snappy, but it is a bit buggy. Goodtop says it's MaxLinear, but I honestly can't tell. It doesn't look like any existing web interface I've seen here though. Nope, I'm not prying off the heatsink. UI is showing a temperature sensor at 48℃, firmware version 1.0.0.6, and hardware version A0.

 

[blunden]

The GT-ST015 is still in the box, but I've configured the ZX310S-8T2XS and it's wall mounted live in my LAN now. Running two interface LACP aggregation to my existing 1G switch, and moved a couple devices over to 2.5G now. The web interface is surprisingly snappy, but it is a bit buggy. Goodtop says it's MaxLinear, but I honestly can't tell.

Yes, the ZX310S-8T2XS may very well be based on a MaxLinear switch chip. We've seen those 8 + 2 port switches using such a chip before.

If I were to guess, the GT-ST015 is likely based on the RTL8372 with one of its two 10 Gbit/s lanes being used as a 2.5GBASE-T port. It's a bit of a waste in my opinion, but I guess people sometimes need that extra 2.5GBASE-T port amd don't want to buy an 8 port switch instead.

 

[jauling]

Yes, the ZX310S-8T2XS may very well be based on a MaxLinear switch chip. We've seen those 8 + 2 port switches using such a chip before.

If I were to guess, the GT-ST015 is likely based on the RTL8372 with one of its two 10 Gbit/s lanes being used as a 2.5GBASE-T port. It's a bit of a waste in my opinion, but I guess people sometimes need that extra 2.5GBASE-T port amd don't want to buy an 8 port switch instead.

I started looking at the cheap-switch-exporter code for Prometheus to use for this switch. The port statistics are totally different, but at least the output is in JSON. With some help from ChatGPT, I've adapted a new authentication mechanism for this model, and now need to translate the /port_statistics.json bit into a /metrics exporter for Prometheus. Quite new to all this, but I have some basic understanding of what's going on

I think I'll have something to release in a few days, depending how lazy I am.

The /port_statistics.json looks like this if anyone cares:

  "PortNum": "10",
  "Port_1": {
    "Port_Id": "1",
    "Port_Status": "Enabled",
    "Link_Status": "1000MbpsFull",
    "TxGoodPkt": "45801815",
    "TxBadPkt": "0",
    "RxGoodPkt": "404721800",
    "RxBadPkt": "412"
  },
  "Port_3": {
    "Port_Id": "3",
    "Port_Status": "Enabled",
    "Link_Status": "2500MbpsFull",
    "TxGoodPkt": "519212529",
    "TxBadPkt": "0",
    "RxGoodPkt": "500901767",
    "RxBadPkt": "0"
  },
...
}

It seems a session cookie only lasts less than 5minutes on this switch, so I do a forced re-login every 4 minutes. What is extremely annoying is that the switch only allows one session at a time. This means if I try to login through my web browser, then my looping exporter session is invalidated. Hopefully once I properly collect the metrics I need in Prometheus/Grafana, then I only will need to browser login if I need to make a change.

 

[ms264556]

It seems a session cookie only lasts less than 5minutes on this switch, so I do a forced re-login every 4 minutes. What is extremely annoying is that the switch only allows one session at a time. This means if I try to login through my web browser, then my looping exporter session is invalidated. Hopefully once I properly collect the metrics I need in Prometheus/Grafana, then I only will need to browser login if I need to make a change.

I work around expiring sessions like this by adding retry logic to the request method. Then I can re-login on failure and try the request again. Depending on the complexity of the request method, I'm either doing the login-retry with a while loop, or by recursively calling the request method (with e.g. an isRetrying param so only one login is attempted).

The advantage is that if later firmware fixes the timeouts then I don't need to update the code.

 

[DaveLTX]

Received both switches yesterday. Quite acceptable shipping wait, approximately a week. The GT-ST015 is still in the box, but I've configured the ZX310S-8T2XS and it's wall mounted live in my LAN now. Running two interface LACP aggregation to my existing 1G switch, and moved a couple devices over to 2.5G now. The web interface is surprisingly snappy, but it is a bit buggy. Goodtop says it's MaxLinear, but I honestly can't tell. It doesn't look like any existing web interface I've seen here though. Nope, I'm not prying off the heatsink. UI is showing a temperature sensor at 48℃, firmware version 1.0.0.6, and hardware version A0.


View attachment 47786

Realtek uses 2 chips to do 8x 2.5G + 2x SFP+

This is definitely maxlinear, very undersized heatsink though IMO

 

[Shonk]

Received both switches yesterday. Quite acceptable shipping wait, approximately a week. The GT-ST015 is still in the box, but I've configured the ZX310S-8T2XS and it's wall mounted live in my LAN now. Running two interface LACP aggregation to my existing 1G switch, and moved a couple devices over to 2.5G now. The web interface is surprisingly snappy, but it is a bit buggy. Goodtop says it's MaxLinear, but I honestly can't tell. It doesn't look like any existing web interface I've seen here though. Nope, I'm not prying off the heatsink. UI is showing a temperature sensor at 48℃, firmware version 1.0.0.6, and hardware version A0.


View attachment 47786

I also have a GoodTop ZX310S-8T2XS (MXL86282S Managed) coming
to replace my XikeStor SKS1200-8GPY2XF (MXL86282C Unmanaged)

Both are the same soc but one has the management interface disabled

I fancied a managed variant anyway i just poped open the Xikestor the heatsink is 72x72x16
its crazy how small it is on the GoodTop i will have to replace that somehow

 

[DaveLTX]

I also have a GoodTop ZX310S-8T2XS (MXL86282S Managed) coming
to replace my XikeStor SKS1200-8GPY2XF (MXL86282C Unmanaged)

Both are the same soc but one has the management interface disabled

I fancied a managed variant anyway i just poped open the Xikestor the heatsink is 72x72x16
its crazy how small it is on the GoodTop i will have to replace that somehow

View attachment 47942

From my experience, that is more a suitable heatsink. No doubt the goodtop is going to have issues

 

[Shonk]

From my experience, that is more a suitable heatsink. No doubt the goodtop is going to have issues

I have alleviated it for now

Dropped under SOC PCB pad thickness from 6mm to 2mm by adding 4mm of copper with MX4 then 2mm of pad to the PCB
Added 1 10x10x10mm heatsink + 2 x 10x10x5mm to the side of the heatsink then Contacted the heatsink to the top of the Case

Its fine now but when i can be bothered to order it i will get a 90x90x15mm + 50x50x15mm heatsink from aliexpress
cut it to as big as i can get to fit maybe 60x90x15?? I will measure when i do it
and drill some pushpin holes for the 4 holes in the PCB that shoud be able to mount a heatsink with pushpins
though the spacing is odd it should work fine

I dont really need to do it anymore but im a tad ocd





I have ordered a 90x90x15 and will cut it to size and drill some holes for pushpins
I have also orderdered a 50x50x15 as a backup plan incase i run into issues

 

[jauling]

Been playing a bit more with my ZX310S-8T2XS. Not sure if the vertical wall mount is negatively impacting the temperatures, but I'm averaging around 53ºC. Seems a bit hot eh? My network closet isn't very ventilated, so it's mostly my own fault, but maybe the heatsink is enough.

With the help of DeepSeek, I created a web exporter for prometheus, not sure if I'll release the code as I feel it's pretty janky. The temperature is exposed in /status.json and looks like this:

{
  "temperature": "53",
  "sys_ipv4": "10.9.9.9",
  "sys_ipv6": "2001:db8::1",
  "sys_ipv6_ll": "fe80::1e2a:ffff:ffff:ffff",
  "sys_macaddr": "1C:2A:A3:EE:EE:EE",
  "fw_ver": "1.0.0.6",
  "hw_ver": "A0",
  "des": "sw2",
  "modle": "ZX310S-8T2XS"
}

In somewhat related news, I asked Goodtop if this switch would ever get SNMP(v2), and they flat out told me it's too wimpy for that feature. So maybe it's not a MXL86282S limitation, but something else.

 

[Shonk]

I removed the GoodTop ZX310S-8T2XS heatsink today
it was 19x19x16mm and was only 5g in weight
the 10x10x10mm i had stuck to the side of it was 6g
and the 2 x 10x10x5mm i had stuck on the other side where 3g each
so its not only small but there's also no weight in it

I was going to use a heat gun set at 80c to soften up the epoxy or glue
anyway i put it in the freezer for 4.5 hours at -29c in a ziplock bag

I managed to get the heatsink off in the end but it was hard work

It turns out its applied with heat conductive silicone e.g. rubbery and i would have been far better off using heat to make it easier

you just turn the heatsink left and right dont pull as you may damage the solder joints pulling
and dont over do it take your time

The max heatsink size you can fit is 70x50 or 70x49 if you dont want to use the few mm past the pcb
and maybe 18mm height

Here's some images i presume all hongrui glued heatsinks use this stuff

 

[Shonk]

My heatsink and pushpins havnt come yet
but wanted to get that glue off

I have temp stacked a pile of copper on there with just MX4 paste inbetween layers
sitting there like a jenga tower

1 x 2.0mm Copper M.2 Heatsink
3 x 1.0mm Copper M.2 Heatsink
6 x 0.8mm Copper M.2 Heatsink
1 x 5.0mm Aluminium M.2 Heatsink

Its a bit slippy slidey but will be ok for a few days
doesnt even do that bad considering there is no pressure at all and the layer transfer wont be great between 11 layers

Will edit this with images of the final heatsink when i do it in a few days

It used to be in the mid 50's before i ever opened it



First Poweron


A Hour later

 

[sharkydog]

Hello everybody, just registered in the forum as I am another one hunting for a cheap 2.5g switch and bit the bullet with one ZX310S-8T2XS.
I was skeptical since I did not find any proper documentation or firmware updates prior to purchase, but decided to go ahead anyway as there isn't anything close to this price.

I want to share my thoughts on VLANs and see if I am correct.

access port - set to some bridge id other than 0 in port vlan and add an entry in tagged vlan for the VID (External VLAN) and bridge id from port vlan.
trunk/hybrid port - leave it unconfigured in port vlan (bridge 0) and add VIDs in tagged vlan with bridge ids for the access ports.
As I understand it, bridge 0 can forward to other bridges if there is a matching vlan, but the other bridges can only forward between their ports and bridge 0 ports with matching vlan.

Currently it's on the testing stand, I'll do some testing with separate dhcp server and a laptop to see what happens before I deploy it in the cabinet.

Am I close? Is this crap actually usable for some simple vlan setup (no more than 10ish vlans on 3-4 trunks)? Or I should not waste my time and produce some more e-waste?

 

[Shonk]

I think i got a bit carried away with project that heatsink is too small

I Ended up using a shim to raise the heatsink over all of the components apart from one that i had to trim a touch off the corner of the heatsink
with a flap pad
It should also help with thermals due to the mass of copper


2 x 77x22x2MM M.2 Copper Shim Under SOC 30g x 2 = 60g
8 x 15x15x2MM Copper Shim Under SFP+ 4g x 8 = 32g
1 x 77x22x1MM M.2 Copper Shim Ontop of SOC = 15g
2 x 77x22x2MM M.2 Copper Shim O Ontop SFP+ Cages 30gx 2 = 60g
2 x 77x22x5MM M.2 Aluminium Heatsink Ontop of SFP+Cages 8g x 2 = 16g
1 x 90x50x15 Heatsink = 81g
The 50x50x15 Heatsink i had as a backup was 36g btw i didnt end up using it

 

[Shonk]

I Dumped the XikeStor SKS1200-8GPY2XF MXL86282C SPI earlier
and checked the soc and such anyway it has a nice heatsink
but a 1.5mm pad between it so put a shim in there with MX-7 while i had it apart
and did a bit undeneath also

 

[jauling]

I experienced my first full switch failure of my ZX310S-8T2XS yesterday. I got alerted that the switch wasn't pingable, then I noticed web interface was also unreachable, and devices local on the switch also couldn't talk to each other, and the LAG to my Dell switch was also not passing traffic. Device was showing lots of blinky lights though. I resorted to yanking and reseating the power, and the switch came back to life almost immediately. I don't really suspect my Dell switch to be the issue, as it also has another LAG that goes to my NAS, that has been solid.

The only logs I have that are relevant are from my uplink Dell switch, which has been showing that this GoodTop switch has been quite unstable since I put it to work. This is just the flapping I saw since I was on vacation on the 28th of April:

Informational    3 - Informational    RAM    06/04/2026 18:54:48    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    06/04/2026 18:54:48    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    06/04/2026 18:54:48    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    06/04/2026 18:54:44    %LINK-I-Up: gi1/0/15
Informational    3 - Informational    RAM    06/04/2026 18:54:44    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    06/04/2026 18:54:25    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    06/04/2026 18:54:25    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    06/04/2026 18:28:47    %LINK-W-Down: Po2
Major    2 - Major    RAM    06/04/2026 18:28:47    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Major    2 - Major    RAM    06/04/2026 18:28:41    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    05/04/2026 08:06:12    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    05/04/2026 08:06:09    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    05/04/2026 08:06:06    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    05/04/2026 08:06:06    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    04/04/2026 06:59:11    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    04/04/2026 06:59:08    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    04/04/2026 06:59:05    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    04/04/2026 06:59:05    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    02/04/2026 17:32:56    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    02/04/2026 17:32:52    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    02/04/2026 17:32:49    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    02/04/2026 17:32:49    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    02/04/2026 07:01:53    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    02/04/2026 07:01:50    %LINK-I-Up: gi1/0/15
Major    2 - Major    RAM    02/04/2026 07:01:47    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    02/04/2026 07:01:47    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Informational    3 - Informational    RAM    01/04/2026 20:45:13    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    01/04/2026 20:45:13    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    01/04/2026 20:45:13    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    01/04/2026 20:45:08    %LINK-I-Up: gi1/0/15
Informational    3 - Informational    RAM    01/04/2026 20:45:08    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    01/04/2026 20:44:58    %LINK-W-Down: Po2
Major    2 - Major    RAM    01/04/2026 20:44:58    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    01/04/2026 20:44:58    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Major    2 - Major    RAM    01/04/2026 20:44:58    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    01/04/2026 20:44:58    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    31/03/2026 07:29:11    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    31/03/2026 07:29:11    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    31/03/2026 07:29:11    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    31/03/2026 07:29:07    %LINK-I-Up: gi1/0/15
Informational    3 - Informational    RAM    31/03/2026 07:29:07    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    31/03/2026 07:28:58    %LINK-W-Down: Po2
Major    2 - Major    RAM    31/03/2026 07:28:58    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    31/03/2026 07:28:58    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Major    2 - Major    RAM    31/03/2026 07:28:58    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    31/03/2026 07:28:58    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    31/03/2026 06:06:42    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    31/03/2026 06:06:42    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    31/03/2026 06:06:42    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    31/03/2026 06:06:38    %LINK-I-Up: gi1/0/16
Informational    3 - Informational    RAM    31/03/2026 06:06:38    %LINK-I-Up: gi1/0/15
Major    2 - Major    RAM    31/03/2026 06:06:28    %LINK-W-Down: Po2
Major    2 - Major    RAM    31/03/2026 06:06:28    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    31/03/2026 06:06:28    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Major    2 - Major    RAM    31/03/2026 06:06:28    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    31/03/2026 06:06:28    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    30/03/2026 20:11:12    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    30/03/2026 20:11:12    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    30/03/2026 20:11:12    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    30/03/2026 20:11:08    %LINK-I-Up: gi1/0/16
Informational    3 - Informational    RAM    30/03/2026 20:11:07    %LINK-I-Up: gi1/0/15
Major    2 - Major    RAM    30/03/2026 20:10:58    %LINK-W-Down: Po2
Major    2 - Major    RAM    30/03/2026 20:10:58    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    30/03/2026 20:10:58    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Major    2 - Major    RAM    30/03/2026 20:10:58    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    30/03/2026 20:10:58    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Informational    3 - Informational    RAM    30/03/2026 09:00:11    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    30/03/2026 09:00:11    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    30/03/2026 09:00:11    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    30/03/2026 09:00:07    %LINK-I-Up: gi1/0/15
Informational    3 - Informational    RAM    30/03/2026 09:00:07    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    30/03/2026 08:59:58    %LINK-W-Down: Po2
Major    2 - Major    RAM    30/03/2026 08:59:58    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    30/03/2026 08:59:58    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2
Major    2 - Major    RAM    30/03/2026 08:59:57    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    30/03/2026 08:59:57    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    28/03/2026 09:17:54    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    28/03/2026 09:17:51    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    28/03/2026 09:17:48    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    28/03/2026 09:17:48    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Informational    3 - Informational    RAM    28/03/2026 09:14:42    %LINK-I-Up: Po2
Informational    3 - Informational    RAM    28/03/2026 09:14:42    %TRUNK-I-PORTADDED: Port gi1/0/16 added to Po2
Informational    3 - Informational    RAM    28/03/2026 09:14:42    %TRUNK-I-PORTADDED: Port gi1/0/15 added to Po2
Informational    3 - Informational    RAM    28/03/2026 09:14:38    %LINK-I-Up: gi1/0/15
Informational    3 - Informational    RAM    28/03/2026 09:14:37    %LINK-I-Up: gi1/0/16
Major    2 - Major    RAM    28/03/2026 09:14:27    %LINK-W-Down: Po2
Major    2 - Major    RAM    28/03/2026 09:14:27    %LINK-W-Down: gi1/0/16
Major    2 - Major    RAM    28/03/2026 09:14:27    %TRUNK-W-PORTREMOVED: Port gi1/0/16 removed from Po2
Major    2 - Major    RAM    28/03/2026 09:14:27    %LINK-W-Down: gi1/0/15
Major    2 - Major    RAM    28/03/2026 09:14:27    %TRUNK-W-PORTREMOVED: Port gi1/0/15 removed from Po2

I've sent GoodTop support these logs, as well as my /status.json output today, but I really have no idea what kind of response I'll get.

 

[jauling]

I haven't gotten anywhere with GoodTop support, but my upstream managed Dell switch is showing LAG flapping again. I also noticed devices on the GoodTop switch also had link flapping yesterday and today, so I suspect the GoodTop switch might lock up again pretty soon, like last time. GoodTop support says the temperatures I'm seeing, around 55-58℃ is normal. I suspect the switch is rebooting during these link down events, since the port statistics packet counts are quite low today.

Anyone else experiencing any sort of instability with this (janky) GoodTop switch?

 

[Shonk]

Havnt had any problems at all it hasnt been rebooted since i put it back after fitting the heatsinks about 2 weeks ago

I did say that heatsink is a waste of space and far too small

 

[jauling]

GoodTop is recommending I try a 12V2A power adapter. I really don't think that's the issue. Your heatsink replacement process seems like a bit more than I'm willing to undertake, but I may consider at least finding a slightly bigger one than the original.