TL-SG3210XHP-M2 - [Port]/6/Reload port Tw1/0/x phy firmware.
Hello,
I'm a new owner of a TL-SG3210XHP-M2 for about a month now, all looked good until today the switch had to be fully powered off and back on, upon powering back up, my sfp+ ports came back on but none of my POE+ ports, that was on FW 1.0.3, tried to unplug/replugged/change ports, force interface up by CLI or GUI... to no avail...Link was always down. All seemed to be working in the last few days, but don't think I tried to fully unplugged the switch and back on, on reboot by CLI or GUI.
My OC200 and AP's are still getting POE+ power but no link on FW 1.0.3.
On FW 1.0.3 after restoring power to the switch
#show interface status
Port Status Speed Duplex FlowCtrl Active-Medium
---- ------ ----- ------ -------- -------------
Tw1/0/1 LinkDown N/A N/A N/A Copper ------> OC200
Tw1/0/2 LinkDown N/A N/A N/A Copper ------> AP1
Tw1/0/3 LinkDown N/A N/A N/A Copper ------> AP2
Tw1/0/4 LinkDown N/A N/A N/A Copper ------> PC for testing
Tw1/0/5 LinkDown N/A N/A N/A Copper
Tw1/0/6 LinkDown N/A N/A N/A Copper
Tw1/0/7 LinkDown N/A N/A N/A Copper
Tw1/0/8 LinkDown N/A N/A N/A Copper
Te1/0/9 LinkUp 10G Full Disable Fiber
Te1/0/10 LinkUp 10G Full Disable Fiber
Only thing that seemed to work was to restore FW 1.0.2, weird?
System Description - JetStream 8-Port 2.5GBASE-T and 2-Port 10GE SFP+ L2+ Managed Switch with 8-Port PoE+
Hardware Version - TL-SG3210XHP-M2 1.0
Software Version - 1.0.2 Build 20210201 Rel.60966
Bootloader Version - TP-LINK BOOTUTIL(v1.0.0)
TL-SG3210XHP-M2#show image-info
Image Info:
Current Startup Image - Exist & OK
Image Name - image1.bin
Flash Version - 1.3.0
Software Version - 1.0.2
Next Startup Image - Exist & OK
Image Name - image1.bin
Flash Version - 1.3.0
Software Version - 1.0.2
Backup Image - Exist & OK
Image Name - image2.bin
Flash Version - 1.3.0
Software Version - 1.0.3
POE+ works on FW 1.0.2 but getting these logs repeatably, ports 6 to 8 have nothing plugged in.
I've opened a SR with TP-Link, awaiting a response.
show logging buffer
#2021-12-06 20:30:22,[Port]/6/Reload port Tw1/0/6 phy firmware.
#2021-12-06 20:30:08,[Port]/6/Reload port Tw1/0/8 phy firmware.
#2021-12-06 20:29:54,[Port]/6/Reload port Tw1/0/7 phy firmware.
#2021-12-06 20:29:40,[Port]/6/Reload port Tw1/0/6 phy firmware.
#2006-01-01 08:03:21,[Port]/6/Reload port Tw1/0/7 phy firmware.
#2006-01-01 08:03:08,[Port]/6/Reload port Tw1/0/6 phy firmware.
#2006-01-01 08:02:57,[Port]/6/Reload port Tw1/0/5 phy firmware.
#2006-01-01 08:02:49,[Link]/5/Tw1/0/4 changed state to up.
#2006-01-01 08:02:46,[Port]/6/Reload port Tw1/0/4 phy firmware.
#2006-01-01 08:02:40,[Link]/5/Tw1/0/3 changed state to up.
#2006-01-01 08:02:35,[Port]/6/Reload port Tw1/0/3 phy firmware.
#2006-01-01 08:02:29,[Link]/5/Tw1/0/2 changed state to up.
#2006-01-01 08:02:24,[Port]/6/Reload port Tw1/0/2 phy firmware.
#2006-01-01 08:02:15,[Link]/5/Tw1/0/1 changed state to up.
#2006-01-01 08:02:13,[Port]/6/Reload port Tw1/0/1 phy firmware.
#2006-01-01 08:01:58,[Port]/6/Reload port Tw1/0/8 phy firmware.
#2006-01-01 08:01:45,[Port]/6/Reload port Tw1/0/7 phy firmware.
#2006-01-01 08:01:32,[Port]/6/Reload port Tw1/0/6 phy firmware.
#2006-01-01 08:01:05,[Port]/6/Reload port Tw1/0/8 phy firmware.
- Copy Link
- Subscribe
- Bookmark
- Report Inappropriate Content
"I think it is indeed something with the 2.5G chipset/phy, it impacts just those ports, and in teh logs you can see it trying to init those pports over and over in a loop"
Do you know if a new version of the HW has a different chipset ?
I have the impression the field of 2,5 gbps POE switches if very small - because of lack of proper chipsets ???
- Copy Link
- Report Inappropriate Content
- Copy Link
- Report Inappropriate Content
WESPL wrote
@IctEss My first TL-SG3210XHP-M2 on fw version. 1.6 hours 1.0 worked until the end of December 2022, between Christmas and the new year it was damaged.. just like a few other people.. I am not a supporter of conspiracy theories, but ... a new device returned to me around May 2023 but in the old version 1.0. This time, just in case, I cut all traffic between the controller and the Omada cloud, leaving only local service (I use a VPN). Christmas has passed, and the new year (the period when the first device stopped working). It's January - it's working for now. But in the meantime, after 3-4 months of operation, I sent back two other small unmanaged 5-port 2.5G TL-SG105-M2 for warranty ... could it be a problem with the chipset that supports these 2.5G ports? Any faulty batch since it affects devices with 2.5G ports? Regarding the question of an alternative, there is not much of it - I used a Zyxel 8x1G + 2x2.5G + 2xSFP+ switch for the time of the TL-SG3210XHP-M2 failure.
What's your experience with the Zyxel ?
How did you manage the mix of brands ?
Since my issues I have been looking at their products as well, also considering their 10G offering (technology's a year later since my last decision so) ...
- Copy Link
- Report Inappropriate Content
- Copy Link
- Report Inappropriate Content
wojo wrote
Have a 2.6 model now, hoping that one stays stable.
Still have a 1.6 in the basement I expect to fail one day on a power cycle.
Have you tried removing the power from the 2.6 model and see how it behaves ?
Do we know what the fundamental difference is between v 1.x and 2.x ?
From reading in this community it seems 2.x is causing less issues, but are they completely gone ?
(that would make an even bigger case for insisting on a v 2.x replacement model)
I'm also wondering about this :
The first one I had had to be switched off (meaning : power removed) , and that didn't result in faulty ports.
The problem with the first unit and the replacement unit after that, were caused after its UPS providing power ran out of battery power.
The UPS'es were of different make, but they had one thing in common : neither of them had double-conversion technology (they were line-interactive).
It all just might be coincidence, but I'm curious if there's something about this that might cause (slightly-)above normal peaks / phase shifts that are problematic for the chipset in the switch ...
- Copy Link
- Report Inappropriate Content
@IctEss
Both of mine failed just unplugging them as I was reworking my server closet.
I should follow through on the RMAs now that v2.6 exists.
- Copy Link
- Report Inappropriate Content
@WESPL I also think it has to do with the 2.5 gbps chipsets .
Looking at the current marketplace, there is very little movement - hardly any decent (managed, at least 8 ports) 2.5gbps switches, let alone POE+(+) support ...
I suspect there will be no big 2.5gbps wave, and vendors will just skip to 10gbps.
If only it wouldn't take so much time to get some maturity price/feature wise ...
- Copy Link
- Report Inappropriate Content
- Copy Link
- Report Inappropriate Content
@IctEss There's no power switch and no 'clean shutdown' in software. So it's not like I could have done anything differently.
- Copy Link
- Report Inappropriate Content
@IctEss Looks like it has MARVELL AQR415C-B0-C flip-chip BGA package.
- Copy Link
- Report Inappropriate Content
Information
Helpful: 1
Views: 15063
Replies: 80