High CPU and log filling with " kernel: vtss_phy_wait..."

[Samgoober]

I have a WISP Switch, 8 port, AC 150 watt switch. I noticed the CPU is runing constantly at >90%, and the log is packed with the following:

Oct 28 11:09:42 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 3
Oct 28 11:09:42 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 4
Oct 28 11:09:43 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 5
Oct 28 11:09:44 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 6
Oct 28 11:09:45 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 0
Oct 28 11:09:45 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 1
Oct 28 11:09:46 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 2
Oct 28 11:09:46 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 3
Oct 28 11:09:47 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 4
Oct 28 11:09:47 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 5
Oct 28 11:09:48 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 6
Oct 28 11:09:49 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 0
Oct 28 11:09:50 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 1
Oct 28 11:09:50 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 2
Oct 28 11:09:51 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 3
Oct 28 11:09:51 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 4
Oct 28 11:09:52 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 5
Oct 28 11:09:53 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 6
Oct 28 11:09:54 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 0
Oct 28 11:09:54 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 1
Oct 28 11:09:55 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 2
Oct 28 11:09:56 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 3
Oct 28 11:09:56 kernel: vtss_phy_wait_for_micro_complete:155 MCB timeout, port_no 4

I dropped to the cmdline and ran top. "poe thread" is at the top of the CPU, taking most of the CPU (>85%). The firmware is 1.42, Board Rev F.

The switch as two AF-5x in it, four Rocket-M2, and is daisy-chained to another switch just like it. I'm not finding anything on the forums or Google that might offer a clue as to what it causing this. The switch uptime is a little over five days. I'm fairly sure the CPU may be due to the switch having to log all these errors/warnings. I turned off logging to see if that would help, but it didn't. (I know that wouldn't fix the underlying cause but thought it may help with the CPU...) Any thoughts?

Thanks
Sam

[sirhc]

Yea, your running code with known bugs like this.

I send out PM BUG ALERTS, I sometimes wonder if people ever read them?

This is what I would do:
1) Disable SMTP alerts if enabled
2) Reboot the switch
3) Upgrade to v1.4.5rc7
4) Re-Enable SMTP alerts if you had them on.

If you still have a problem after installing the latest version which right now is v1.4.5rc7 then report the issue.

Anytime you ever run into a problem always load the LATEST code then if the problem persists write up a post.

[Samgoober]

I figured, but didn't want to upgrade right away without checking first. I didn't see this bug listed anywhere so I was trying to be careful. I'll proceed with updating it. Thanks for getting back to me so quickly.

Sam

[sirhc]

I figured, but didn't want to upgrade right away without checking first. I didn't see this bug listed anywhere so I was trying to be careful. I'll proceed with updating it. Thanks for getting back to me so quickly.

Sam

There are several bugs that can affect CPU usage that were fixed, one is SMTP alerts, one is Discovery protocols.

[sirhc]

If the problem still persists after upgrade please swap out switch and if problem goes away when swapped then please RMA that unit.

[Samgoober]

I upgraded the firmware on the switch. It rebooted, then wouldn't come back up. It continually reboots itself now. I will ship it back to you soon as I can get out there to pick it up.