XYMer's Home away from Home

When http://bbs.xlr8yourmac.com is down (i.e. always)
It is currently Thu Nov 23, 2017 9:36 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 116 posts ]  Go to page 1, 2, 3, 4, 5  Next
Author Message
PostPosted: Sat Mar 11, 2017 4:17 pm 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
All of a sudden the RT-N12 with Shibby Tomato will no longer boot. After powering down and back up, it just sits there with all the lights continuously on. Seemingly dead. What to do to try to fix? Anything short of completely resetting and reflashing as per viewtopic.php?f=17&t=6775&hilit=Shibby&start=50#p83716? 30-30-30 reset, or what?

In the meantime, I took out the N-10P, which I had completely configured back in 2013, had working with basically the same config as the N12, then put away for a rainy day. Was working fine before I put it back in its box. Now, thinking that I could just use one of the recent N-12 config backups on it, I put it in service, but I can find no way of logging into it at 192.168.1.1. Managed to get something going from Network prefs setup--no idea really what it is--with Verizon DNS. Getting a weird IP now of "pool-xxxxxxxxxx.nycmny.fios.verizon.net." Not something I want to stay with for long. Wondering even If I'm behind the router with that kind of address. Maybe it's completely bypassing the NAT.

So now a two part nightmare: the N-12 dead, at least for now (possible cause: had a power failure last night, but was working OK until this happened today mid-afternoon), and second part, not being able to get into the N-10 at 192.168.1.1, to restore the config, which I would keep using until I can figure out what's going on with the N-12, and possibly get that going again. Tried clearing its NVRAM from the WPS button, but that didn't make any difference.

Tried telnetting into it, pinging, no dice. Also smb no dice. Zero idea what to try next. Looks like it just doesn't want to respond at that IP. Assuming it's not something else going on with the N-10, if I could only get into that and restore the N-12 config, I'd have something working again the way it should.


Top
 Profile  
 
PostPosted: Sat Mar 11, 2017 8:09 pm 
Offline
Benevolent Dictator
User avatar

Joined: Mon Apr 21, 2008 2:03 am
Posts: 14666
So, how exactly are you connecting here?


Top
 Profile  
 
PostPosted: Sat Mar 11, 2017 8:16 pm 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Not sure what I'm doing anymore. Been at this all day. Had Diagnostics in Network do the configuration.


Top
 Profile  
 
PostPosted: Sat Mar 11, 2017 8:58 pm 
Offline
Benevolent Dictator
User avatar

Joined: Mon Apr 21, 2008 2:03 am
Posts: 14666
I figured you're flustered.

So in Sys Prefs>Network, it says the Router IS 198.168.1.1?


Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 4:35 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
BDAqua wrote:
I figured you're flustered.

So in Sys Prefs>Network, it says the Router IS 198.168.1.1?

Nope, but should be 192.168.1.1. "Router" there is showing a version of the IP, so it appears that the router (whatever that can mean now) is at Verizon????? Plus, no idea what pool<IP>.nycmny.fios.verizon.net is. What is a "pool" IP address, and why am I getting one of those? And now this is what shows up in Terminal:

Last login: Sun Mar 12 08:32:29 on ttys000
pool-xxx-xx-xx-xxx:~ xxxxx$


Before this happened, I had the DNS for all the clients configured to the router at 192.168.1.1. And now, I can't have more than one client up and working at the same time. If I do I get "Another device is using...."

Could probably solve all this in a matter of minutes if I could get into the damn router and restore one of the configs I was using.


Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 7:19 am 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
You cannot restore configs across routers like you're proposing. The config only works for the unit you backed up the config from. The config will restore things like the router's interface MAC addresses, which are unique per router, and often contain model-specific settings that don't work on other models.

I would turn off the power switch on the N12, press and hold the recessed reset button, then turn on power, while still holding reset. Keep holding reset until the power light starts very slowly flashing on and off. If it starts flashing you're in very good shape, the router is in recovery mode. If you assign 192.168.1.2 to your ethernet card you can visit the recovery website at http://192.168.1.1 and re-flash the firmware onto the router and/or reset nvram (preferred order is to reset nvram then upload a new firmware). So long as you restore the same version of the flash as you put on it last time you should be OK to restore the config. Shibby has been making a lot of nvram changes lately so restoring an old config will be problematic to a non-starter, but if its the same version it'll work fine.

As for the N10, you can try resetting the nvram. With it off press and hold the WPS button, turn it on and keep holding down the button, it should almost immediately start quickly flashing the power light. Let it sit there for a second or two then release, flashing should stop. Router will then boot normally for a first boot, which is slightly longer than an already-configured boot. Assuming its shibby the DHCP server should be on by default so it should at least give out IPs and basically work.

I've had to reset four N12s in the past couple weeks due to power issues. Three of the four would broadcast the WiFi network but not allow any devices to connect to it, the fourth allowed devices to connect but ran glacially slowly. At a site 10 miles away a good 10-12 N12s are still trundling away, but that site has fewer power events than the main campus (they seem to wait until closing time to start screwing with our power, I kid you not - 15 minutes after close the events start and continue until about 30 minutes before we open).

BTW, assign a 192.168.1.2 address to your NIC and start pinging 192.168.1.1 - at two points during the device booting it will respond to pings with a ttl=100 for several seconds. If it hits ttl=100 but never gets around to responding to pings later with a normal ttl then it's at least booting off the CFE, which is where the recovery website lives.


Last edited by MonkeyBoy on Sun Mar 12, 2017 7:36 am, edited 1 time in total.

Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 7:33 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
A million thanks MB for the help. Will see if I can get the N-12 going again, with those instructions.

In the meantime, after 4 hours sleep + being in a state of near total exhaustion, I finally had a huge WTF/aaaaargh!!! moment this morning. Yesterday, I had disconnected all the cables, and totally fuzzyheaded, and frenzied plugged the WAN into the wrong port of the N-!0. Now it's in the right port and the N-10 is up and working just fine. Can get into it now at 192.168.1.1.

Already discovered that I couldn't use the backup from the N-12, as it refused to do that when I tried. Found a backup I did only once after setting up the N10 and restored that + manually reentered the OpenDNS and Dnsmasq entries from screenshots and notes I had taken, so that's OK too. Huge relief. It's always the stupidest things done in a frenzy.

Now that the N-10 is doing OK, I will take a breather from all this router insanity.


Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 7:38 am 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
OH! If you're in the recovery mode only port 1 will work for the recovery website. They added this restriction a while back, I always forget about it because the older routers didn't care.

It also doesn't stay in recovery mode for long, a couple minutes at most, so you've got to stay on top of this and not flake out and do something else (something I'm guilty of, but to be fair it's usually because 3 or 4 people suddenly decided to call/walk into my office like they just realized I existed and can dump all the problems they've been saving up for weeks onto me).

BTW, I stumbled across your post because I remoted into work to take a look at something else, but as part of confirming some of this stuff in my own bleary eyed Sunday morning state I discovered Comcast has dropped the internet connection on the line they just installed last week. Why? Can't tell. Just doesn't give the router an IP. Great. Just great. God I hate Comcast.


Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 8:01 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Nothing like trying to focus in a state of near total exhaustion. One interesting thing I discovered is that now that I'm back on OpenDNS, I don't get quite the Internet speed I was getting with the VZ DNS/servers.

When I was on the VZ DNS, from the configuration done by Diagnostics, wonder what that IP address, and router IP with pool<IP>.nycmny.fios.verizon.net + in Terminal, Last login: Sun Mar 12 08:32:29 on ttys000
pool-xxx-xx-xx-xxx:~ xxxxx$
was all about????


Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 2:48 pm 
Offline
Benevolent Dictator
User avatar

Joined: Mon Apr 21, 2008 2:03 am
Posts: 14666
Quote:
Yesterday, I had disconnected all the cables, and totally fuzzyheaded, and frenzied plugged the WAN into the wrong port of the N-!0

Arrgh, I should've got that, I did that years ago with the same symptoms. :(


Top
 Profile  
 
PostPosted: Sun Mar 12, 2017 4:17 pm 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
What was fuzzyheaded was that in panic mode, I diconnected all the cables without first noting where they came out (last I touched this router was in December 2013.) But I was clearheaded enough to get out the instruction manual to see where the cables should go back in. Only problem with that was that in the manual the cable positions are shown completely reversed.


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 4:58 am 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
Since it sounds like you plugged a LAN port of the N10 into the router, you probably were given a public IP address, or at least an address that Verizon could provide the reverse DNS name to. The N10 was basically a switch, albeit one with a DHCP server attached that would occasionally have given you an IP that wouldn't be able to go anywhere (since the WAN port wasn't attached to anything).


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 6:14 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Yeah, certain that I plugged a LAN port into the ONT (the Fios modem.) Your explanation--though I don't completely understand it--of how I was nevertheless able to get an IP sounds like that's what happened.
Quote:
The N10 was basically a switch, albeit one with a DHCP server attached that would occasionally have given you an IP that wouldn't be able to go anywhere (since the WAN port wasn't attached to anything).

This part I don't get, since it didn't just "occasionally" give me an IP, I was able to get online pretty consistently with that "pool IP nycmny.fios.verizon.net" address (and it stayed the same, depending which was connected, the Mini or the iMac--they got different IPs) whatever that was, as long as the Mini or the iMac weren't both connected at the same time. When I search for nycmny.fios.verizon.net, I get hits re. MCI.

In Network, "router," the internal IP, was showing the beginning of that same public IP range, and I was able to ping it. Instead of the final three digits of that IP, it was using just the first of those 3 digits. For example, if it was 71.68.229.141, the internal IP or "router" was 71.68.229.1 Maybe it would have been possible to extrapolate the internal IP from that. Maybe nothing, but I wasn't happy with the security implications of that.

And I have a question about the power outage I suffered the night before this happened: since it's normal practice to power cycle the router simply by unplugging its power--something I've done any number of times--why would a power outage have been any different? Why would the router have noticed the difference and reacted badly?

Still have some serious first aid to do on the N-12, but the N-10 is doing just fine, and the wireless signal it's giving for the iPad and Apple TV is quite strong enough--the Mini and iMac are hardwired.

And now we may get up to 24" of snow, just when I was ready to put my winter coat away.


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 8:16 am 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
Because you plugged a LAN port into the LAN port you had two DHCP servers, each giving out different information.

The problem isn't a power outage per-se. It's when power drops, then comes back, then drops, then comes back, in just the right lengths to corrupt the NVRAM. I've only had it explained to me so I don't precisely understand why it happens (the boot process on these routers is best understood by starting them up with a serial cable attached, and I've never cared enough to solder a serial port onto one), I just have lousy enough power at the main site to know that it certainly happens.

I had my oil changed a couple weeks ago and the guys at the shop were saying I needed to swap out my snow tires because it had been warm for a while. Man I wish it worked that way.

I would hope that putting the N12 into rescue mode will get it back to life. You may even be able to get it to work just by wiping nvram by holding down wps for a couple seconds while its booting.


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 8:27 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Thanks for the explanation. Tried clearing the NVRAM with the WPS button, but might try again.

Getting ready to do first aid on the N-12, but no longer have the Shibby I originally downloaded (must have been saved on the G3, which I completely wiped last year), and the link you originally gave for that, back in December 2013, http://tomato.groov.pl/download/K26RT-N ... %20RT-Nxx/ gives a not found now.

This is what I originally used from your post back then:

shibby tomato-K26-1.28.RT-N5x-MIPSR2-115-MiniIPv6.trx and shibby tomato-K26-1.28.RT-N5x-MIPSR2-115-Max.trx,

I can only find later versions from 2014. Everything appears to be the same for the Max and Mini versions, except -121- instead of -115-

Second and third ones down on the page. Not even certain these are Shibby:

http://tomato.groov.pl/download/K26RT-N ... %20RT-Nxx/

If these won't work with the config backups I have, maybe you know where else to find them?


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 8:45 am 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
You should be able to use 132 and earlier, that's when the big changes happened, but it's possible you may need to set things up from scratch.

According to the list here:
http://tomato.groov.pl/download/K26RT-N/

It looks like 120 is the oldest he still has online. I would probably opt for 132 if I didn't have the previous version, since either way you could have to set it up from scratch, but may as well grab 120 and 132 and hope for the best. The builds you want will be listed as

tomato-K26-1.28.RT-N5x-MIPSR2-###-Max.trx
tomato-K26-1.28.RT-N5x-MIPSR2-###-MiniIPv6.trx

Where ### is the build (120 or 132).

I normally run Max on my N12s but both should work. Any of the K26 non-USB builds will work, really. Its just a measure of what services are available to be turned on in the router's firmware - by default the extra stuff is left off.


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 10:25 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Thanks for that. Suddenly remembered that I had put the G3 backup away in a drawer. Opened up that ancient external and found the original firmware that I used back then. Moved it to the Mini, where I'll be doing this. So, at least that's taken care of.

Btw, what would doing this from scratch mean? I'm supposing that the only way would be from recovery mode. Or could that mean a factory reset, using the stock Asus firmware and then flashing with the Shibby, which is how I did this from scratch?


Top
 Profile  
 
PostPosted: Mon Mar 13, 2017 10:37 am 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
From scratch means not using a configuration file, it's going through the website and adjusting settings to however you want them to be from their defaults.


Top
 Profile  
 
PostPosted: Tue Mar 14, 2017 1:52 pm 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Think it's probably toast. Tried maybe 5X, won't enter recovery mode, power light never starts blinking. Tried 30-30-30 reset, no dice. Cleared the NVRAM from the WPS button, nothing happens. Anything left to try?


Top
 Profile  
 
PostPosted: Tue Mar 14, 2017 2:03 pm 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
Damn that sucks. The CFE loads really early in the boot process. Cable it up with 192.168.1.2 configured on your ethernet card, start pinging 192.168.1.1, then turn it on. Does it ever respond? It should respond very quickly, go silent for a while, come back with ttl 100 for about 5 seconds, then go silent again until the firmware boots. If you get ttl 100 you could tftp a small firmware over to it like a ninja.

If it's completely dead and never responds to pings then something bad happened, probably power related. Resetting the firmware holding the WPS button at power on at least rules out corrupt NVRAM. Brownouts and spikes and similar events can damage equipment. The way to really be safe from them is to run off something like a Smart-UPS that runs off the battery 100% of the time, but its crazy to buy a $250 UPS for a $30 router.


Top
 Profile  
 
PostPosted: Tue Mar 14, 2017 2:15 pm 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Well, with the rebate and low price at Newegg at the time, both the N-10 and N-12 together ending up costing me a massive $20. So I think I got my money's worth. And very glad that I went on a shopping spree back then and got both of them. First the N-10, then learned they were giving away the N-12 for nothing also. At least I had the N-10 as a fallback. And if it's still available, I might get another one--seems to do the job just as well as the N-12, for my purposes.

Don't see why pinging it will do anything, since it just never boots. But if some miracle happens, please explain how to tftp a small firmware. What firmware, and how to tftp?

Oh, and to make this N-10 last as long as possible, should I trust the Tomato Tx power default of 0, use the 42mw that comes up first (for this router, is the hardware default the same as 42mw?) Or should I use an even lower value, like 10mw to keep it running as cool as possible, as long as I keep getting an acceptable signal at the two wireless devices?


Top
 Profile  
 
PostPosted: Tue Mar 14, 2017 2:40 pm 
Offline
User avatar

Joined: Thu May 15, 2008 8:13 pm
Posts: 9568
Location: Caught between the moon and NYC
These routers first load the CFE and then they load the firmware. If there's a problem with the firmware it will still respond to pings when the CFE loads. The N12s I worked on the past couple weeks all had corrupted their firmwares but would load the CFE when I cabled up to port one with 192.168.1.2 on the NIC. Can't leave it at DHCP, there's no DHCP server this early in the router's boot process. I think the command from terminal is tftp but my Mac's powered off in the bag and I'm about to pack it in for the day. Go home, crawl into bed, and try to get some sleep.

All routers default to 42 as far as Tomato is concerned. I would change it to zero, which defaults to whatever value the driver thinks should be the value. You might want to enable interference mitigation under Advanced -> Wireless. I've pretty good luck with WLAN Auto with Noise Reduction or Non-WLAN depending on the environment. Non-WLAN is if you have non-WiFi wireless devices within range causing interference, while the other is if most the worst offenders are all WiFi. Unfortunately without a signal analyzer you can't even see non-WiFi signals, and they're damned expensive, so basically try one and then the other over the span of a few days and try to gauge which works better in your environment.

In my experience the N10 is noticeably slower than the N12 but I have a LOT of WiFi clients trying to connect to the thing so it might simply be overwhelmed with connection attempts and responding to beacons and who knows what else.


Top
 Profile  
 
PostPosted: Wed Mar 15, 2017 5:51 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
Well, wasn't ready to do a firmware tftp--did some research and that looks rather complicated. But before even getting into that, tried pinging the router at 192.168.1.1 to see if it would respond in any way. Set a location in sys prefs Network to 192.168.1.2, trying these two options: "DHCP from manual address" and "Manually" (is either of those the correct way to do this?)

Cabled to what might be port 1, but also tried 2 and 3. Tried starting pinging both before and after powering on the router. Everything I tried resulted in "No route to host."


Top
 Profile  
 
PostPosted: Wed Mar 15, 2017 7:03 am 
Offline
Benevolent Dictator
User avatar

Joined: Mon Apr 21, 2008 2:03 am
Posts: 14666
How's that storm treating you?


Top
 Profile  
 
PostPosted: Wed Mar 15, 2017 7:46 am 
Offline

Joined: Sat Apr 11, 2009 2:15 pm
Posts: 5978
Location: NYC
We got far less than expected here, only around 5" or so. But it's a nasty, sludgy, icy 5". However, the night before the storm arrived I discovered leaking coolant from my Civic just after warming it up for a while. Hoping it's only minor, like a hose or loose hose clamp and not the radiator. Filled up the reservoir, which was empty--the radiator was still full--but will have to do some shoveling later this afternoon before I can get to opening the hood to see if it kept on leaking without the car running--which would be very bad news indeed. It's been in the 20's at night, and not good if there's no antifreeze. Bad timing all around. Even if there were no snow under the car and I could get underneath, I wouldn't be able to see anything, as the bottom area under the radiator and its connections is covered by a plastic cowl. Won't drive it until I can bring it in.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 116 posts ]  Go to page 1, 2, 3, 4, 5  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group