Question WLAN issues and checklist.

[Scav]

Hi guys,

I should already know this. I really should. I went to school for exactly this sort of felgercarb.

The problem is (and there's no guarantee this laundry list is set up in any particular logical order):

* Canon MG2500 is setup with a static IP on the wifi network. Is talking properly with the local access point. IP address for the printer has been set to 192.168.1.9, with a subnet mask of 255.255.255.224 (I only have, and expect to have a handful of devices on this network, so I deliberately set a tight subnet for this network). Local access point has a MAC address reservation for this printer and CAN ping said printer locally.

* Neither the laptop, nor my PC connected to the access point can resolve the printer's IP. Neither devices can see the printer. Neither devices can ping each other.

* Both laptop and PC are running Windows 7, can ping the gateway, can ping localhost, and the Canon drivers are up to date. The entire setup was working just last week, and apparently decided to stop working the week I needed to print my W2's for tax time.

A google crawl has been . . . interesting, but ultimately unhelpful yet. Besides . . . I know you people.

So! Any ideas on anything I missed describing this problem?

 

[Quick_Nick]

Worth a shot if you haven't already tried:

Windows can be a bit like that, it's hit and miss with finding my network printer too. If it doesn't find the printer, you have to manually add the printer's IP address, etc. in the Add Printer wizard.

 

[Scav]

Worth a shot if you haven't already tried:

Mmm, I did think about that, and no joy on that even now -- it seems the third-party communication between PC->access point->printer is the bottleneck; I'm 90% sure of that from what I'm (not) seeing.

Additional information I should've included in the first post: The access point I'm using is a Cisco E1000 router.

---------- Post added at 05:46 AM ---------- Previous post was at 04:52 AM ----------

Oh, for the love of cheese.

I just contacted one of my instructors. He suggested I adjust the subnet mask to 255.255.255.0.

Once I did that and refreshed the router, one of my wife's print jobs immediately cleared the spooler and hit the printer. Moreover, my PC can now see the same printer. :facepalm:

SO!

Lesson learned: too tight of a subnet mask is too secure of a system; can cause complication with third-party devices.

Thanks, QuickNick, for trying to assist!

 

[Urwumpe]

Yes, the subnet mask is a critical aspect... it is also used for broadcasting messages, which are important for many systems.

Also, you are likely using a home network (192.168.xxx.xxx for example) anyway, so no need for a tight subnet mask anyway. Better choose a good random number for the third byte.

And with IPV6, all of this subnetmasking and other historic stuff is obsolete.

 

[garyw]

* Canon MG2500 is setup with a static IP on the wifi network. Is talking properly with the local access point. IP address for the printer has been set to 192.168.1.9, with a subnet mask of 255.255.255.224 (I only have, and expect to have a handful of devices on this network, so I deliberately set a tight subnet for this network). Local access point has a MAC address reservation for this printer and CAN ping said printer locally.

As 192.168.Y.Z is a class C private managed network there is no need to set a tight subnet mask.

* Neither the laptop, nor my PC connected to the access point can resolve the printer's IP. Neither devices can see the printer. Neither devices can ping each other.

Do you have DNS configured or is it attempting to resolve via broadcasts/hosts? Are all the IP's static or is there DHCP in there?

Does windows have it's firewall running? What are the ip's and subnet masks assigned to all the devices?

And with IPV6, all of this subnetmasking and other historic stuff is obsolete.

True but let's be honest here, IPv6 Looks like a MAC address got drunk and threw up.

 

[Urwumpe]

True but let's be honest here, IPv6 Looks like a MAC address got drunk and threw up.

Yes. But I can now assign a IPv6 address in my home network to every hair on my head. Isn't that great? :lol:

 

[Scav]

As 192.168.Y.Z is a class C private managed network there is no need to set a tight subnet mask.

True, in retrospect. When I configured banjaxed planned did what I did, I did it in the mind of practicing security, and in my mind I wanted a short limit of what could and could not access the wifi in the event of a WPA breach. Live and learn, right?

Do you have DNS configured or is it attempting to resolve via broadcasts/hosts? Are all the IP's static or is there DHCP in there?

DNS appears to be running at its default, and I do have DHCP passing out addresses on demand. Everything seems to be running under the SN:AFU principle now.

Does windows have it's firewall running? What are the ip's and subnet masks assigned to all the devices?

That's an interesting thing, because that was one of the things I checked. It didn't make a difference with or without the W7 firewall on my end, so I was able to conclude that the bottleneck was at a point other than either my wife's laptop or my PC. The only other logical point of failure was either the WAP or the printer itself, and the printer was less likely due to several perceived factors (component age was one of them). And a botched subnet wasn't really in the cards; the printer was the only peripheral (well, that and my iphone, but I rarely use that to connect to the wifi here) device connected to the network, and I had all of the computers on the same subnet anyway. It all worked until it decided not to. The part that confuses me is why everything worked for so long, and then decided to take a nap abruptly like that. What changed? :hmm:

True but let's be honest here, IPv6 Looks like a MAC address got drunk and threw up.

:rofl:

Can I use that in normal (yeah, I know. When do geeks ever have normal conversations?) conversation?

Yes. But I can now assign a IPv6 address in my home network to every hair on my head. Isn't that great? :lol:

If your hair can communicate over the internet . . . then I'd consider it a very eventful day in München.

---------- Post added at 10:15 AM ---------- Previous post was at 09:54 AM ----------

And now I have more practice for my standard duties as an admin: printer upkeep. Jammed infeed in the cartridge. Time to pull out all the paper and run the roller cleaning cycle. :woohoo:

 

[Ripley]

Yes. But I can now assign a IPv6 address in my home network to every hair on my head. Isn't that great? :lol:

...Depends...

Sorry! :rofl: Couldn't resist...

 

[Linguofreak]

Yes. But I can now assign a IPv6 address in my home network to every hair on my head. Isn't that great? :lol:

More than likely, you can assign an IPv6 address on your home network to every hair that has ever existed on any human body.

 

[Quick_Nick]

From random Google numbers, it appears you could at LEAST assign an address to every atom in all 7 billion people alive.

 

[Izack]

I keep forgetting how delightfully overkill IPv6 is. Makes the protracted overhaul time that much more bearable because at least this time, we'll never have to do it again (yes I'm aware of the hopelessness of that optimism.)

Now if only it would also follow that public static IP adresses become easier to obtain / retain without dishing out too much. But I'm not counting on it.

 

[Linguofreak]

From random Google numbers, it appears you could at LEAST assign an address to every atom in all 7 billion people alive.

That would be from the entire IPv6 address space (the numbers I'm seeing say you'd need about 123 bits to address every atom in the set of living humans, out of the 128 bits available, so you could do it 32 times over). Urwumpe was talking about the number of addresses you'd have in your home network under IPv6 (which will be at least a /64 block but could be as big as a /48, depending on the generosity of your ISP). While the number of atoms in every living human wouldn't at all fit into this, the number of hairs on every living human (7 billion people with 5 million follicles each) would fit comfortably into a /64 (it would take 55 bits). In fact, the number of hairs that each person now alive will have produced by the time they die (assuming a hair follicle produces a hair every 4 years, estimated based on this , and assuming a life expectancy of 80 years) still fits in a 64 (60 bits). And, taking the largest estimate I've found for the number of people (116 billion) who have ever lived (for the sake of squeezing the address space as hard as possible), the combined lifetime production of individual hairs produced by every human that has ever lived so far still fits within a /64 (though it does require all 64 bits, taking up 63 percent of the block).