Belkin RT3200/Linksys E8450 WiFi AX discussion

This sounds like a cache issue. LuCI Attended Sysupgrade grabs https://sysupgrade.openwrt.org/json/v1/overview.json, looks at the values in the latest entry near the top and produces a list of appropriate candidates from it. Since the ASU server updated yesterday morning, I see 23.05.3 as available in the same place you're seeing the .2 version. (This all happens starting at https://github.com/openwrt/luci/blob/master/applications/luci-app-attendedsysupgrade/htdocs/luci-static/resources/view/attendedsysupgrade/overview.js#L443 )

First try flushing your browser cache (ctrl-shift-R or whatever on that page). If that doesn't work, maybe try logging out of LuCI and logging into the router again.

3 Likes

Screenshot 2024-04-02 200525
Clearing cache google chrome helped. Thank you.

2 Likes

Mine is also 48SAR601.0GA and no issues so far. I have it since November 2022 and been running OpenWrt since.

sw_version=1.0.0.1
modelNumber=E8450
cert_region=US
hw_revision=1
hw_version=48SAR601.0GA
manufacturer_date=2020/09/18

The RT3200's LED is indeed white rather than blue.

2 Likes

@daniel Are you still collecting logs from bl2-for-debug-snand-issue.bin?

Here's mine:

1 Like

Official wiki now updated after following through the steps myself.

https://openwrt.org/toh/linksys/e8450#recovery_from_openwrt_kiss_of_death_okd

Let me know of any errors.

@FanOfOpenWRT this should answer your question about what dongle you need for a recovery.

14 Likes

Note: All my devices mentioned below are in a different physical location (and different country) so I am unable to debug by connecting serial or UART or JTAG etc.

RT3200-1 is still chugging along as main router.

Both RT3200-2 and RT3200-3 (both just dumb APs) did not come back to life on reboot or just power ON at normal room temperature. Both these RT3200's have been (temporarily) replaced with two TP-Link Archer A7 v5 (ath79, ath9k, ath10k) as dumb APs.

I am hoping that the RT3200-1 (main router) continues to keep working. If the RT3200-1 also fails in the future, then I will have to repurpose one of the Archer A7's as the main router and lose one dumb AP.

I hope that this whole issue is fixed in a (hopefully very near) future OpenWrt SNAPSHOT and / or U-Boot and / or ARM Trusted Firmware release.

I currently have four RT3200 routers that have all been running with 6.1 kernels for some time.

I just compiled the latest snapshot, installed it on one of the routers and it rebooted and worked perfectly. Then I proceeded to install the exact same firmware on a second router and that router didn't boot back up. I appear to be in the situation where I see no power light, which suggests I probably need to dig out the serial cable etc.

Unfortunately I don't have time to look into this at this moment, so I will need to do it a week on Friday, but there evidently is still some sort of intermittent problem when re-flashing the firmware.

When I do get to the point of opening the case up and connecting a serial cable is there anything that anybody would like me to do to look for problems?

1 Like

Can anyone help out with:

Even though we don’t know what causes OKD, I really think listing the possible causes of OKD might be a good place to start.

And working through the ARM Trusted Firmware A v2.9(release) changelog to see if anything looks relevant also seemed worthwhile.

I am trying to see if I can help motivate more coordinated effort in trying to tackle OKD! It seems like rather a challenging puzzle that might benefit from collective effort.

My present thinking, is OKD is caused by ARM Trusted Firmware A v2.9(release). Am I wrong? Shoot me down and we’ll move onto the next item on our list. Once we have a list.

2 Likes

Perhaps Daniel still needs more data (beats me), but you can collect the debug logs before you fix with the bl2-for-debug-snand-issue.bin file

Without a scientific evidence, I blame attended system upgrade. That's the only differenece I see so far from my dead router and the one still alive. Both 6.1.
And to make it more complex both survived 2 asu, until one fall.

2 Likes

Not much to go by, but the one time mine failed was when attempting upgrade using attended system upgrade.

I did not use attended system upgrade and was running 23.05.2 when mine failed to boot. Dont think that is a factor.

if we look at this epidemiologically -
only routers on openwrt would have their health status reported on this forum.
there is not a standard way of sampling the health of the population of routers.
we have a cluster of failures reported recently (2 months?) tfrom the openwrt subpopulation.

we don't know how many 'stock' devices are failing because their caregivers don't report here.
we don't know how many of those failed with upgrades from linksys/belkin.
we don't know how many routers failed with the initial atempts at conversion to openwrt/ubi

we do know that a handful? (10? less than 20?) people have reported errors recently.
i know that i started responding to this thread because of those error reports.

i seriously wonder if the device sufferes from poor design. starting with power supply; 2A seems underrated for the potential peak demand. that plus some capacitor/ tolerance issues, a boot loader that requires rereads of the same blocks, and maybe some mis-specified bootloader driving power settings.

bottom line is we don't know the extent of the problem - the unknown unknowns.
that should not stop, but will add uncetainty to, a search for solutions.

i love this opernwrt community.

4 Likes

With regards to OKD. I have never experienced one. I installed firmware v1.0.0 and never updated.
My network gear is plugged into a UPS and has lost power only once unexpectedly (18 hour power outage). I briefly ran (a month maybe) with a min freq of 600Mhz but have gone back to the default 437Mhz.

I am pretty hard on this router. I compile from 23.05-SNAPSHOT regularly (whenever I see a kernel or driver update). I have tweaked busybox to allow writing to flash (ssh history). I had a couple custom scripts that wrote to flash on reboot. I also run this as my router with Cake, so it gets a work out.

Not one hiccup. Best OpenWrt device to date. Evidence of one.

I still do wonder about the power. Daniel says he hammers this model for testing too, but uses a more hearty power supply, not the factory provided one.

On this theory alone I'd be willing to buy a new power supply/cable and try it on mine just for S's & G's. What is a good suggested 3rd party power cable for the RT3200 (or what exact specs should I seek/exceed)?

1 Like

I've been using the factory power supplies, cheap PoE extractors, and other power options that I think would be equally as likely to cause the issue and I haven't had a single occurrence of the problem either. I have one unit rigged up on the bench as it were, also trying all of the things people say has caused the error with the hopes of getting it to fail, but it has yet to fail. I've monitored power consumption, and I have yet to get any of this model router to draw anywhere near the rated 1A from the factory supplies delivered with the RT3200s. So far, it still looks like a random sampling of errors with no viable leads.

2 Likes

Guess it’s my turn to unfortunately jump on this band wagon.

E8450 running 23.05.2 and initial OWRT from installer 1.0.3. Had been working flawlessly. Turned device off as I’m moving house. Device now no longer appears to turn on, no LEDs, no signs of life. Unfortunately I don’t have any time right now to investigate due to the chaos of unpacking a house of things :cry:

2 Likes

12V-2A gives us 24W, which is more than the entire board needs, if external USB device is not used. The extra juice is really only needed to drive external USB devices. I think the USB port only supply up to 500mA, so it really only uses 2.5W max.

I would think the mt7622 SoC would consume not more than 5W at full tilt, leaving plenty for the rest.

IMHO, the stock power adapter is adequate.

3 Likes

Has anyone encountered the dreaded OKD with installer v1.0.2 or earlier?

3 Likes