IPQ40xx (EA3500) borked by Attended 25.12.2

Performed Attended SysUpgrade to address the CVE related issues associated with 24.10.

EA8300 = no problemo, worked flawlessly

EA3500 = She’s a Brick … router

Troubleshooting:

  1. Power-cycle 3* to get alt partition to boot. No joy. Have not had that occur before. Hmm. No SSIDs and no hardline handshaking. Try again, and again. Nada
  2. Press and hold reset button for 10+ seconds to clear settings. Set fixed IP of 192.168.1.3 on PC and try hardline connection again. Nada.
  3. Power down router and upon power up press WPS button to get into failsafe. Nada. Just never-ending rapid flashing power light.
  4. Started investigating TFTP method but info seemed a bit vague on the details that “I” would need to complete such a task.
  5. Switch to Linux laptop and repeat all of the above. Nada
  6. After about 5 hours a WiFi sweep detected a SSID from the router that I was able to make a connection. It was the previous 24.10 firmware so the alt-boot did finally occur.

Observations: The Attended SysUpgrade selected 25.12.2 but that didn’t work out so well (see above). By contrast, the EA3500 product page lists 24.10.4 as the Current version despite the releases of 24.10.5 and 24.10.6. The site’s Firmware Selector agrees with SysUpgrade’s choice of 25.12.2. However, the Table of Hardware agrees with the EA3500 product page’s choice of 24.10.4. So the score is kinda 2:2 but my single piece of experiential evidence has its thumb on the 24.10 side of the scale.

Unfortunately, the router’s System Log does not capture anything following my initial login to perform the Attended SysUpgrade. There is an 8-hour block missing. I surmise that the block is when the unit was failing to boot 25.12.2.

Wed Apr 1 07:07:48 2026 daemon.err uhttpd[1668]: [info] luci: accepted login on / for root from 192.168.1.11
Wed Apr 1 15:56:50 2026 cron.err crond[1594]: time disparity of 528 minutes detected
Wed Apr 1 15:59:18 2026 daemon.err uhttpd[1668]: [info] luci: accepted login on / for root from 192.168.1.23

Just sharing but always appreciate new learning if you have something to share about this.

Thx

ToH is manually updated, so often lags behind on what's available. If it's shown in the Firmware Selector, then it was built successfully and is available for the displayed version.

Did that device get any config updates over the 24->25 upgrade? Might be something changed and you need to upgrade without saving the config...

Yes, was on 23.05 and have been successfully utilizing Attended for the past several updates. Is there a way to get the other firmware’s log file? I’d like to see what’s recorded there that covers the missing 8-hour block that might help explain just what went wrong with the upgrade.

Extremely unlikely, as default logs are in RAM or /tmp (which is also RAM), so evaporate on reboot.

That’s too bad because I think there is more going on here than I’d first thought. Later today, I thought the ISP was having issues but turns out the EA8300 (main router which had accepted 25.12.2) had stopped working. Power cycled it in place but still no joy. Connected it directly to PC and found that it had reset itself to previous FW 24.10.5 but was corrupted. Imported the last 24.10 config and all is good on the EA8300 again. The question remains; why would 25.12 fail and as a result have to fallback to the previous firmware? There’s a nightly reboot cronjob but router was working this morning so something else is causing issue(s). Other than the OW upgrade there have been no other network changes.

If you have an external drive, say a USB stick, you could mount that and set system log_file to somewhere on it and possibly capture whatever is going wrong.

https://openwrt.org/docs/guide-user/base-system/system_configuration?s[]=cronloglevel#system_section

Ok, did that now. Seems to be logging to the USB alright. Waiting for an issue to appear.