Beneficial to do factory reset after firmware update?

[PunchCardBoss]

The settings file is not a database

Sure looks like a flat file to me (which is a Db). Now, maybe your definition of a "simple dump of all the variables" is the same as my definition of a flat file. And that is okay.

In my world, a data dump is one way (export) The fact that you can upload settings implies data structure (Db).

Agreed that it contains user settings, and operational settings that are specific to the FW. And it is also very obvious that the FW performs
1. record updates (such as FW version),
2. transitions from old field names and/or data to new fields names and/or data
3. adds new and/or deletes fields during a FW update/power cycle, some of which may be user or operational.
4. And, I do see records that appear to be operational like WIFI channel.

And, it may indeed have dynamic data (although I would think such data would be better kept in a temp table in RAM.

So far, I have examined the flat files of (3) FW settings. Each is different in the number of records. And each has the FW version in the data set. I will admit, that surprised me. But what it does show is that a FW update includes translation. And by translation, I mean the ability of the FW to modify user/operational settings data that are proper for a prior FW version and modify them. This translation can include new records and fields and/or re-named records and fields.

I am most anxious to see what happens in this next update. I intend to save a snap-shot of my settings just prior to applying new FW. Then another snap-shot directly after FW update (no power cycle). And, finally, another snap-show after power cycling.

I may even try re-loading my first settings snap-shot (under the older FW) and then downloading settings to see if they changed. My hunch is they will not change until I do a power cycle. But that is just my guess.

So what is the take away? Here is my working hypotheses...
1. Settings translation is happening either at time of FW update or power-cycle. I'm not sure which at this time. My guess is that the FW update process itself performs settings translation.
2. A soft-reset does retaining (translation) of login capability (I think this is true from reading other's posts. Please correct if I am wrong). By this I mean login credentials are maintained, but all other user and operational fields go to defaults defined by the FW.
3. IF (and it is reasonable to make this a big IF), settings translation goes well, there is no need to perform a soft reset. I also concede that performing a soft reset after FW update is indeed the "best practice" but not always necessary or practical for those that have lots of user settings to adjust. The issue with this take-away is you just don't know if the translation went well.
4. Asus most often is going to recommend a soft reset because Asus has no control whether the FW update (with settings translation) will work if the existing FW is old(er).
5. The chances of a FW update going well without a soft reset and from the previous FW version is reasonably good (not fool-proof). If performance is off and/or lots of errors in the logs, then reset.

 

[ColinTaylor]

@PunchCardBoss

1. Obsolete variables from old firmware versions are migrated to their replacements in the current firmware each time the router boots up. The process can be seen in the source code.
2. Both soft and hard resets reset all the user settings to their default values including erasing the login account. When you connect to the router you will be back at the initial setup page.
3. As I said in post #3 there is usually no reason to do a reset after a firmware upgrade.
4. If you contact Asus with problem a reset will be the first thing they ask you to try. This is to establish a known baseline and is no different than every other software product.
5. Same as answer 3.

 

[PunchCardBoss]

Both soft and hard resets reset all the user settings to their default values including erasing the login account. When you connect to the router you will be back at the initial setup page.

Ugg. Confirm that. I just read several articles confirming the same.

As I said in post #3 there is usually no reason to do a reset after a firmware upgrade.

Agreed. At least that is what my research is telling me.

Obsolete variables from old firmware versions are migrated to their replacements in the current firmware each time the router boots up. The process can be seen in the source code.

Question1: Does the router go through a "router boot" as the final part of a FW update? My guess is YES. And if so, is there any advantage to power cycling after a FW update?

BTW. Tks for your reply and clarifications - good stuff.

Question2: Then why does this issue keep coming up in this forum - in different threads? I have seen this debate on and off for months. And most replies are short, and without serious explanation. This question is the only reason I stepped in - to try to bring some common sense logic and reason to the issue. And I was prepared to do extensive testing to try to figure it out. Yet, the question stands. Why the constant debate?

 

[L&LD]

While the 'by-the-book' remarks above are 100% true to how a router should react to an update/upgrade, I've seen many instances where that was simply not true for specific routers/networks. Great if it works for you. Not so great if it didn't and you don't even know about some of the tips/tricks to try to get your network reliable and stable again. Including how to do a proper reset to factory defaults.

Q1: Anytime I make a change to the router's configuration, I always do an (additional) reboot. It has saved many headaches since I started doing that 3-minute 'step', sooner, rather than later. Further, I've seen better results too, if the reboot is done 15 to 60 minutes after an update has been completed. I believe this gives the hardware a chance to finalize configuring the software side of things first. If a simple reboot (or two) doesn't work (always being initially performed via the GUI), I then pull the power plug from both the router and the AC wall outlet and let it sit for 2 to 10 minutes. If this doesn't work too, I will then reflash the same firmware I just upgraded the router to.

Q2: If the router/network, at this point, doesn't respond how it is expected to respond... A more nuclear approach is needed to ensure that:

• 1) that the router is really being reset (many posts show this isn't always happening, for various reasons).
• 2) that the router is stable after being reset and the hardware 'sound'.

Consumer WiFi routers combine many different systems in a small and efficient package at a low cost for the performance they're capable of. What they're not designed for is the Enterprise level of reliability. That is why the 'additional' steps may help people to at least diagnose (as much as a consumer can) that the fault is hardware, or not). The people that have the ability to delve deep into the innards' of today's routers think that everything is fixable with the tools they possess. And they may be right many times. However, it is usually far easier to simply reset the router, set it up minimally and manually, and get back to a working network in much less time and effort.

Why the constant debate? Because this is the internet and people who like to argue, do.

 

[ColinTaylor]

Question1: Does the router go through a "router boot" as the final part of a FW update? My guess is YES.

Yes it does.

And if so, is there any advantage to power cycling after a FW update?

No, not normally.

Question2: Then why does this issue keep coming up in this forum - in different threads? I have seen this debate on and off for months. And most replies are short, and without serious explanation. This question is the only reason I stepped in - to try to bring some common sense logic and reason to the issue. And I was prepared to do extensive testing to try to figure it out. Yet, the question stands. Why the constant debate?

Because some people don't understand the processes involved and have invented pseudoscience explanations which they promote in these forums. Unfortunately people that are even less knowledgeable then believe what they are told a repeat it to others. Their argument goes like this: My router was behaving badly for reasons I don't understand. I rebooted it but it made no difference. The next day I wore red socks and rebooted and it was working again. Therefore the solution to everyone else's problem is to wait 24 hours, wear red socks and reboot the router.

 

[PunchCardBoss]

Why the constant debate? Because this is the internet and people who like to argue, do.

The next day I wore red socks and rebooted and it was working again. Therefore the solution to everyone else's problem is to wait 24 hours, wear red socks and reboot the router.

Okay, I think I have stopped laughing enough to write this post.

So much wisdom in each of your posts. Thank you both @L&LD and @ColinTaylor. Both (and others) have been very helpful in demystifying this (silly) debate. Everything that you guys have contributed makes perfect sense based on 40+ years of my own electronic systems experience (although routers are new to me). I really appreciate understanding more about the router's "business back-end".

I will sign off for now and remind all....
"Intelligent people make intelligent choices when they have good intelligence."
and
"Listening precedes learning" "Knowledge precedes understanding".

Tks again for your kind patience with a newbie.

 

[Tech9]

Because some people don't understand the processes involved and have invented pseudoscience explanations which they promote in these forums.

Spot on description. It lowers SNB Forum quality a lot. A good place regularly spammed by amateur waste of time guides. They come 5min after a new member registers here.

 

[OzarkEdge]

Spot on description. It lowers SNB Forum quality a lot. A good place regularly spammed by amateur waste of time guides. They come 5min after a new member registers here.

My AiMesh install notes are a 4-year distillation (<10,000 characters) of what I've learned from others here and from my own study and experience using AiMesh. I think they are a very useful baseline reference for anyone unfamilar with installing an ASUS router in their home. I maintain and improve them for my own use and share them for the benefit of other users. Deriding my so-called 'amateur effort' and similar here is a disservice to this community.

OE

 

[L&LD]

As it is a similar disservice to me and my efforts to aggregate all the best practices, I've learned on these forums these past 9 years (and I have read every post since then).

The people I give the links to seem to be appreciative of my assistance. The waste of time is from others whining about them.

If the beginner's guides are too far below your level of expertise, simply ignore them. Easy peasy.

 

[Tech9]

Deriding my so-called 'amateur effort' and similar here is a disservice to this community.

Your notes are fine. The usual red socks approach is coming from a different source. Another significant difference is you don’t call your notes “recommended” or “best”, but just “my notes”.

 

[L&LD]

Missed the mark again. Best Practices, as discovered on these forums. Partial quotes can slant anything.

Yes, I can recommend the best practices contained within my links, because they worked and continue to work for those that need them. Some of the other best practices were told to me directly (years ago), and a few, I discovered on my own (via the forums). Doesn't make them any less valuable to someone who needs their network working again.

A rose by any other name still smells as beautiful.

But please, stop trying to 'cleverly' label something as 'bad', when it isn't.