ASUS RT-AC68U Firmware version 3.0.0.4.385.20253

[Grisu]

stay with 384 for the moment.

 

[gdp416]

I upgraded again to the latest. This time I did a factory reset after the upgrade and the CPU utilization went back to normal.

Edit March 29 - it didn't work for long, see below.

 

[kozman]

I reverted back to 3.0.0.4.385.20252 because the router was just laggy. Everyone should just submit a Feedback response to Asus and maybe they'll fix it soon.

 

[kozman]

I upgraded again to the latest. This time I did a factory reset after the upgrade and the CPU utilization went back to normal.

Hmm, makes me wonder if I should do it. I'm not super worried about Krook vuln and I have a backup of the config. Did you restore from .CFG backup after factory reset?

 

[OzarkEdge]

Hmm, makes me wonder if I should do it. I'm not super worried about Krook vuln and I have a backup of the config. Did you restore from .CFG backup after factory reset?

No, not if you are debugging a problem.

https://www.snbforums.com/threads/faq-nvram-and-factory-default-reset.22822/

OE

 

[Con]

If I wish to go back to the previous firmware, is it ok to just download the previous version and flash the router?

 

[L&LD]

@Con, yes.

If/when you have persistent issues, a full reset/Initialize will be required to get the router back to a good/known state though.

 

[Con]

L&LD thank you for your quick reply. Am I correct that a reset will return the router to a factory default state and would require setup as if it were just out of the box?

 

[L&LD]

@Con, yes.

See the link in my signature below for further information.

 

[kozman]

If I wish to go back to the previous firmware, is it ok to just download the previous version and flash the router?

That's what I did. I manually downloaded and applied the previous firmware and my router is back to normal. 99.9% there's no need to factory reset after firmware upgrade. No one designs a product with that intent built into it or it (or the company) would not be on the market for long. Only in the really rare occasion where things go sideways and even reverting doesn't work is it better to start from scratch. It's just so odd that some folks have no issue after firmware upgrade and some do. I've not factory reset my router since I bought it. All firmware upgrades worked fine save for this one. And to me it's not worth a factory reset to fix the CPU bug.

We all know and love Asus and I bet they're investigating the issue and will push a fix in short order.

 

[OzarkEdge]

That's what I did. I manually downloaded and applied the previous firmware and my router is back to normal. 99.9% there's no need to factory reset after firmware upgrade. No one designs a product with that intent built into it or it (or the company) would not be on the market for long. Only in the really rare occasion where things go sideways and even reverting doesn't work is it better to start from scratch. It's just so odd that some folks have no issue after firmware upgrade and some do. I've not factory reset my router since I bought it. All firmware upgrades worked fine save for this one. And to me it's not worth a factory reset to fix the CPU bug.

We all know and love Asus and I bet they're investigating the issue and will push a fix in short order.

ASUS recommends that you try the Hard Factory Reset before attempting any advanced troubleshooting.

https://www.asus.com/US/support/FAQ/1039074

But I suppose downgrading firmware is not advanced troubleshooting.

OE

 

[kozman]

ASUS recommends that you try the Hard Factory Reset before attempting any advanced troubleshooting.

https://www.asus.com/US/support/FAQ/1039074

But I suppose downgrading firmware is not advanced troubleshooting.

OE

Yes of course. And like Microsoft, if their patch screws something up, uninstalling the patch usually fixes it. Still, this is obviously a bug in the firmware (for some) if reverting fixes it.

 

[OzarkEdge]

Yes of course. And like Microsoft, if their patch screws something up, uninstalling the patch usually fixes it. Still, this is obviously a bug in the firmware (for some) if reverting fixes it.

My point was that Asus recommends to reset before performing advanced troubleshooting. By inference, it is worth the trouble if you are troubleshooting.

OE

 

[tts]

Almost 100% cpu usage after upgrade.

 

[aaronk]

I was also getting 100% cpu usage on 385.20253, mostly by [mtdblock3] according to top. I saw that there was a new directory in /jffs called asd, and removing that directory immediately resolved the problem.

When the asd process restarts, /jffs/asd is recreated and the cpu goes to 100% again. A persistent fix is to use touch /jffs/asd to make a file that blocks creation of the directory.

 

[aietos]

I was also getting 100% cpu usage on 385.20253, mostly by [mtdblock3] according to top. I saw that there was a new directory in /jffs called asd, and removing that directory immediately resolved the problem.

When the asd process restarts, /jffs/asd is recreated and the cpu goes to 100% again. A persistent fix is to use touch /jffs/asd to make a file that blocks creation of the directory.

I'll try your fix. I was able to reduce CPU consumption (stock Asus 3.0.0.4.385_20253) by simply umounting /jffs (somewhat surprised that's possible if it is clearly in use?).

 

[gdp416]

I upgraded again to the latest. This time I did a factory reset after the upgrade and the CPU utilization went back to normal.

Well, the apparent "fix" of a Factory Reset that I thought had worked, didn't work for long as the persistant CPU utilization was back up to >85% when I checked it a day later. I reverted to 3.0.0.4.385.20252 and all is fine CPU wise again. I will try a Hard Factory Reset after the next upgrade, but doubt that will work. If not, I will recomplain on the Feedback to ASUS and try aaronk's workaround.

 

[gdp416]

Well, the apparent "fix" of a Factory Reset that I thought had worked, didn't work for long as the persistant CPU utilization was back up to >85% when I checked it a day later. I reverted to 3.0.0.4.385.20252 and all is fine CPU wise again. I will try a Hard Factory Reset after the next upgrade, but doubt that will work. If not, I will recomplain on the Feedback to ASUS and try aaronk's workaround.

Well, tried the hard factory reset, and it looked okay for a day, then last night it obviously went back up to > 90% utilization no matter what, and is staying up there. As aaronk pointed out, most of the cpu usage is by [mtdblock3]. Submitted another feedback to ASUS.

 

[ForkWNY]

Not that this helps those who run 68U's as their primary routers, but none of my (4) four 68U's running as AiMesh nodes have high CPU usage on the 3.0.0.4.385_20253-gb338b0e firmware release. All four are below 2-4% CPU on average, though the only way I have of knowing CPU usage on any of them is to SSH in and run "top" to see it. From what I've seen of the "high CPU usage" bugs in the past, the httpd/httpds process is the culprit (the process that runs the web UI for the router). The httpds process does not run however on AiMesh nodes for obvious reasons.

 

[kozman]

Not that this helps those who run 68U's as their primary routers, but none of my (4) four 68U's running as AiMesh nodes have high CPU usage on the 3.0.0.4.385_20253-gb338b0e firmware release. All four are below 2-4% CPU on average, though the only way I have of knowing CPU usage on any of them is to SSH in and run "top" to see it. From what I've seen of the "high CPU usage" bugs in the past, the httpd/httpds process is the culprit (the process that runs the web UI for the router). The httpds process does not run however on AiMesh nodes for obvious reasons.

It's still mtdblock3 that's causing it, not httpd.