384.3 for 3200 Series should be pulled...seriously!

[ispgeek]

Ha! the minute I go back to beta 3 the Network Map fails immediately upon any change of configuration, same with wifi, same with reboot. Upgrading back to the Alpha 4 now and lets see what happens.

 

[ispgeek]

And of course sanity returns to the router, I can at least make normal changes now without having to reboot the device. One notable thing when the router enters this state CPU Core 1 and Core 2 alternately and repeatedly peg at 100% in some dual fight for something yet to be determined. Don't piss on me and tell me it's raining there is something seriously messed up here.

 

[ispgeek]

Oh wow I found one change that even fails on this build. Try enabling or disabling the second 5ghz radio, save. Then go back to main page note failed network map and then look up at the pretty cpu core volley......back n forth back n forth......this so far is the only change that I have found that with the Alpha version triggers it.

 

[muffintastic]

@ispgeek - I can relate, also owning the AC3200 when I was previous build 380.69_2 it worked as it should, I too did the same precedures as you, factory reset old build, power recycled and used the recovery utility. The problems I face are as follows:

• nat: apply nat rules (/tmp/nat_rules_eth0_eth0) error! - meaning all of port forwarding rules are borked and lose internet connectivity randomly
• miniupnd breaks
  
• slow sluggish gui
• poor 5GHz wifi coverage

The only the cure is either power down for 5 minutes or hard reset then reinstall the FW again and works again until it breaks. None of these problems were in the previous branch until upgrading. The problem is all you get is generic answers from people whom are not experiencing problems, and yes I've changed all my ethernet cables and makes no difference.

So, you're not alone @ispgeek this build is borked, stable for few days then everything messes up.

 

[ispgeek]

Software initiated reboots still result in dead network map, the cpu core pegged duel and dead wireless. Power cycle fixes it nice.

 

[ispgeek]

@ispgeek - I can relate, also owning the AC3200 when I was previous build 380.69_2 it worked as it should, I too did the same precedures as you, factory reset old build, power recycled and used the recovery utility. The problems I face are as follows:

• nat: apply nat rules (/tmp/nat_rules_eth0_eth0) error! - meaning all of port forwarding rules are borked and loose internet connectivity randomly
• miniupnd breaks
  
• slow sluggish gui
• poor 5GHz wifi coverage

The only the cure is either power down for 5 minutes or hard reset then reinstall the FW again and works again until it breaks. None of these problems were in the previous branch until upgrading. The problem is all you get is generic answers from people whom are not experiencing problems, and yes I've changed all my ethernet cables and makes no difference.

So, you're not alone @ispgeek this build is borked, stable for few days then everything messes up.

But did you change the Ethernet cable lubricant because everybody knows that unlubricated Cat5e will cause a router to fail especially when it's combined with questionable firmware lol. Thank you for reporting your findings as well. I highly recommend that you upgrade to the Alpha (4) version for the 3200 if you haven't already. It will give you some peace and at least allow you to do basic housekeeping functions without constantly having to reboot the monster with a power cycle....don't touch that 2nd 5ghz radio though because 100% sure you will break things instantly.

 

[ispgeek]

Oh and you definitely don't want to bother with a soft reboot (GUI or any software based reset) because that too will result in wasted time and further your pain.

 

[muffintastic]

Alpha (4) version, hows that for you what's your uptime?

 

[ispgeek]

Oh god before somebody actually tries looking for Cat5e lubricant my bad....you can get off Amazon, comes in injector style spray can for $9.95 Prime.


(Just kidding please don't try such stupidity at home folks...no such thing as Cat 5e lubricant or gulp valves (auto humor))

 

[ispgeek]

Alpha (4) version, hows that for you what's your uptime?

Well it was all night until my schedule software reboot kicked in about two hours ago....I have since turned it off lets revisit this question in 24, 36, 48 and 72 unless it fails before then. In any case it fixes the any change to config breaking things that is most definitely in the current release code....that in itself is worth loading.

 

[Dave K]

5 noted experiences of 3200’s running the 384.3 firmware successfully in this thread and 1 having issues. The evidence is, your issues aren't enough evidence to pull the firmware.

My 3200 didnt like 384.3.. I had 2 different threads on issues on 384.3 I went to the Alpha (383.4) and it seems stable.

 

[scjr]

My 3200 didnt like 384.3.. I had 2 different threads on issues on 384.3 I went to the Alpha and it seems stable.

Yes, I remember that. Quite a few 3200’s haven't had issues. I get the ops frustration. He wants it to work. Unfortunately, as we all know, firmware upgrades can be fickle. I’ve had a problem or two upgrading myself. Resetting to factory defaults normally fixes the issues. In this case, seems that’s not helping. I just think pulling the firmware isn’t the answer, unless the developer thinks that’s what needs to be done.

 

[ispgeek]

Yes, I remember that. Quite a few 3200’s haven't had issues. I get the ops frustration. He wants it to work. Unfortunately, as we all know, firmware upgrades can be fickle. I’ve had a problem or two upgrading myself. Resetting to factory defaults normally fixes the issues. In this case, seems that’s not helping. I just think pulling the firmware isn’t the answer, unless the developer thinks that’s what needs to be done.

As with any firmware upgrade there is risk, generally the risk is to the device itself, many of these devices were built with the equiv of $10 in actual parts costs and one of the areas that almost always takes the short end of the stick is nonvol memory and overall design. While the qty is there the quality isn't. Nonvol memory is one of the weakest links in these devices as it has a limited number of read/write cycles and that's it. Depending on the overall design this is often overlooked by engineers adding all the bells and whistles totally forgetting they might be memory intensive thus reducing the anticipated lifespan on the device itself. Consumers expect a five year lifespan, the router industry however builds for three, anything above that is a bonus (and most devices go well beyond it), however, recently we've seen a number of very high end devices plagued by firmware upgrade issues (or just plain firmware issues) but specifically a high number of devices that fail even on a properly applied firmware upgrade and I suspect the reason in part may be that non-vol memory is failing prematurely which would most certainly result in a bricked or partial bricked device and especially bricked so badly there is no recovery. Normally you would only see such failures in aging devices but that isn't the case anymore in fact Netgear is dealing with their own issues on the R9000 as I type this.

You may have noticed that from time to time all manufacturers seem to get a wild hair up their butts and they substantially move memory allocations around making those config backups worthless and causing generally higher than normal failure chaos. Many will give this or that or the other reasons for doing so I suspect their primary reason is to reduce the impact to those areas of memory written often to avoid total failure. Kind of a manual TRIM of sorts that's used in SSD world and Windows much for the very same reasons. Now this can go well or go very badly....if the memory is failing in general then any change at all would likely cause failure. This is only a hunch but its based on a solid foundation on these devices and it would explain the manufacturers forcing the firmware on folks (as they don't want a sudden spike of in warranty repairs as it looks real bad) and the only way to prevent non-vol related ELF is to do exactly what we're seeing with this firmware from the change perspective.

 

[ispgeek]

With that long explanation above the reason I strongly encourage the temporary pulling of the firmware is to prevent others from getting into the same situation and since we don't yet understand what we're dealing with and we are dealing with people far less equipped to deal with it than I am staying on the side of caution will prevent the forums from blowing up with pissed off people who now have a very expensive brick....in todays world one cannot assume people understand the risk they accept when they upgrade firmware because most are clueless even after being warned. So somebody has to be the responsible one and do the right thing to save them from themselves. A temporary hold won't kill anyone but it sure will prevent a lot of headaches and frustration.

 

[scjr]

With that long explanation above the reason I strongly encourage the temporary pulling of the firmware is to prevent others from getting into the same situation and since we don't yet understand what we're dealing with and we are dealing with people far less equipped to deal with it than I am staying on the side of caution will prevent the forums from blowing up with pissed off people who now have a very expensive brick....in todays world one cannot assume people understand the risk they accept when they upgrade firmware because most are clueless even after being warned. So somebody has to be the responsible one and do the right thing to save them from themselves. A temporary hold won't kill anyone but it sure will prevent a lot of headaches and frustration.

Thoughtful response.

All I’m saying is, there will always be a percentage of firmware upgrades that will go haywire. This isn’t reserved for the 3200 only. A percentage of other models will have upgrade issues as well. It’s the nature of firmware upgrades, changes made by the manufacture that cause issues with current settings, etc.

Have you tried 384.4 alpha? Previous poster with 3200 was having issues and moved to that build and it solved his issues.

Hope you can resolve the problems you’re experiencing.

 

[ispgeek]

So here's what we know fails immediately upon upgrading to the release version in some 3200's

• Port forwarding including DMZ (ironic that the router does continue to listen for it's own traffic on port 8080 if enabled but nothing else works)
• Network Map fails immediately upon making any config change until router is fully power cycled
• Wireless both radios fails immediately upon making any config change until router is fully power cycled
• Web History does not populate data
• AI Protection appears to be working but nothing populates over time
• GUI is extremely sluggish or non-responsive
• CPU Utilization between both cores pegs at 100% alternating back n forth constantly
• Software based reboot does not recover the device, the only recovery method is power cycle

Now if you upgrade to the Alpha release the list changes substantially...

• Wireless both radios fails immediately upon making any config change to the second 5ghz radio until router is fully power cycled
• Network Map fails immediately upon making any config change to the second 5ghz radio until router is fully power cycled
• Web History does not populate data
• AI Protection appears to be working but nothing populates over time
• CPU Utilization between both cores pegs at 100% alternating back n forth constantly when any change is made to the second 5ghzadio when everything else fails
• Software based reboot does not recover the device, the only recovery method is power cycle
• Otherwise things work well or much better than with the release version. At least the device functions for its intended purpose.

 

[ispgeek]

Thoughtful response.

All I’m saying is, there will always be a percentage of firmware upgrades that will go haywire. This isn’t reserved for the 3200 only. A percentage of other models will have upgrade issues as well. It’s the nature of firmware upgrades, changes made by the manufacture that cause issues with current settings, etc.

Have you tried 384.4 alpha? Previous poster with 3200 was having issues and moved to that build and it solved his issues.

Hope you can resolve the problems you’re experiencing.

I appreciate your engagement however you just did it too....you aren't reading the responses completely otherwise you would have known I'm already on the alpha and yes there is more than enough smoke to justify a temporary hold on that firmware. When you have issues this serious being raised the smart thing to do is err on the side of caution until the issue is better understood. This costs Merlin nothing to do and the costs he saves others by acting responsibly on his part is huge not to mention his reputation because people will blame his firmware even though the issue really is Asus's. Notice I have not assigned blame other than firmware, Merlin makes mostly cosmetic changes the issues seen here are NOT cosmetic, I suspect these are basecode issues and in all likelihood Asus already knows about them but did what these asshat mfr's are so classic for and gambled that it would not impact the majority of people and I suspect that its going to backfire on them. Netgear is dealing with their own similar issue right now with the r9000 fwiw.

 

[ispgeek]

I could care less about any other models....the only one I'm concerned about is the flavor of firmware for the Asus RT3200.

 

[scjr]

I could care less about any other models....the only one I'm concerned about is the flavor of firmware for the Asus RT3200.

That ends my engagement here. Good luck to you.

 

[ispgeek]

That ends my engagement here. Good luck to you.

I assume that's because you realize the ask on my part is fair as it applies to this specific version of firmware for this model of device only. It would be reckless and entirely inappropriate to expand such a request universally but I have seem the impact the firmware has on this device and its nothing pretty and clearly I am not alone.