You didn't specify what router hardware you bought, so it is impossible to give any specifics. However, there are several generic approaches to resolving your issue(s).

1) Reset the router to its default configuration with all network devices disconnected. Usually, this means holding the reset button for at least 9 seconds. The button is typically located inside a small hole on the back or bottom of the router.

2) Since you are dealing with a router, there will be one network port clearly labeled WAN which MUST be connected to the cable modem.

3) Starting with one PC, reconfigure its network connection to require DHCP and connect it to one of the router's LAN switch ports.

4) Open a command console and type IPCONFIG. Check that IP, gateway and DNS addresses as well as the subnet mask have been assigned by the router's DHCP server. Open a web browser and see if you can access any web site.

5) Repeat steps 3 and 4 for additional PCs.

6) Network printer configurations do not work reliably with DHCP as the IP address may change with a new lease. Pick a fixed IP address that is outside the router's DHCP range (check the manual) and configure the printer's IP address and subnet mask accordingly.

7) On one of the PCs either create a new printer definition or modify an existing one to point at the IP address assigned in step 6.

8) Repeat step 7 for any additional PCs.

You can use DHCP and have it give the same IP address each time to the devices.

As KKU stated, it will be impossible to get real granular with this until we know the exact brand and model of the router.

What you will need is the MAC address of each device. ( do ipconfig /all and note the "Physical Address" of the network adapters. It will be in a xx-xx-xx-xx-xx-xx format) In your router you can assign this MAC address to a specific IP address and it will stick.

For your printer, you will have to print a setup page to determine what the mac is. Once again, without specifics .....

I have the same setup the OP describes but I never did hook up the printer/scanner to the router direct. I have it hooked direct to the main computer which is part of a workgroup. The printer is shared over the network but nobody uses it, the noobs just get me to print the odd doc so I will have to test again make sure it works.

I will follow this thread as I would like to see if hooking the printer/scanner direct to the router is a better option.

My Canon MP830 printer/copier/fax is connected to my file/print server via USB, so I have no personal issues with network printers. However, as a retired Windows/Linux administrator, I have had to deal with many work related network issues including networked printers. Assigning a static IP address to a networked printer that is outside of the router's DHCP pool of addresses is the only sure way that all computers can access it.

There are three idioms that apply, especially with the various unknowns in a forum post:

1) Keep it simple.

2) Read the fine manual.

3) When you find yourself in a hole, put down the shovel.

 

 

Hi Mumble,

 

If you only have a few network devices (ie between 10-15) then it is always easiest to just set up a static IP for each and never have to worry about things again.

 

The important considerations with regard to static addressing are that all addresses remain subnetted to the same network ID (in order to "see" the rest of said network) and retain the correct upstream gateway information so they won't have issues "finding" resources locally or on the internet.

 

The easiest option with a home network is to tailor your IP addressing scheme to match the default of your router.  The first step in switching from a "dynamic" addressing scheme at home to a "static" on is to disable the router's DHCP server (of course understand that anyone coming to your home and wanting to use your internet will then have to assign a static IP that corresponds with your network temporarily while they are connected to your network).

If your router for example uses:

IP               -        192.168.0.1

subnet        -        255.255.255.0   (subnetting in layman's terms tells the protocol which portion of the IP address denotes the host ID and which portion is the network ID)

For all devices connected to the same switch to be able to "see" one another they all must have the SAME subnet information.

 

Then your first network device could use:

IP                     -     192.168.0.2   <--------- change the LAST digit only for additional network devices such as printers, laptops, cellphones, TV's etc. etc.

subnet              -      255.255.255.0

gateway            -      192.168.0.1    <------ the network device's UPSTREAM which in this case is the router so its IP needs to be there

primary DNS      -      192.168.0.1    <------ who the network device needs to "talk" to in order to find things locally or internet (and since the router will have auto-discovered your ISP's DNS for it's own UPSTREAM "talking") so you again enter your router's IP address there.

second DNS       -   LEAVE BLANK   < --------  while you can put one of your ISP's DNS servers in there as a "failover" it is not necessary.

 

^^^ rinse and repeat.

 

done!  

 

The above is the most simplistic way of setting up a STATIC home network.  Also, you can turn those services which you turned off back on.

 

the Monk

 

A picture is worth a thousand words so perhaps this will help.

The only problem I'm having is being able to always and reliably see my two PC's and the Printer under the "Workgroup" path. As I've said I have no issues with either PC accesing the internet or in printing to or scanninbg from. I just can't seem to get my devices to appear by default under Workgroup even though they are always there, they're just not visible until I randomly poked around with the network for a bit.

 

Unfortunately what you're describing is a problem related to the "election process" with regard to the "Computer Browser" service.  Essentially when running a network (workgroup) without the benefit of a server someone still has to sort of hold a "master list" of paths on the network etc.  This is somewhat accomplished by the "Computer Browser" service in a very WINS-like (legacy windows) fashion.

What I'm saying is, since you're not running a server and aren't interested in using static IP's (this means the "computer browser" service will require more updates to keep network paths etc. correct) your best bet is to keep at least one of the machines on your network consistenly powered on and part of the network.  This will ensure that the entire network is "learned" by the service and will be propogated throughout the network to other computers running the same service. 

Of course you can't control exactly when this data is propogated over to the other systems so keeping at least one system on at all times should help with at least that system retaining correct information.

 

EDIT:  Of course it goes without saying that "network discovery" must be allowed on all machines and the actual workgroup name must be the same on all machines.  Oh and under advanced TCP/IP properties you should make sure that "Netbios over TCP/IP" is set to enabled since network discovery will be using that.

Thanks Monk, that actually made some sense to me except that the issue occured only *when* I was using static IP's and not when using DHCP supplied IP's. I'm agnostic between static and dynamic as long as I can get all of the functions I'm looking for which is reliable net access and printer use as well as reliable local Workgroup access.

I did pretty much assume that was the function of the Computer Browser service and generally did have it on on both PC's except for the occassional test to the contrary also both PC's had the same workgroup entitled "WORKGROUP" after all I ran that same workgroup for years when I had DSL as opposed to cable.

Not sure what you mean by Network Discovery though unless you mean Network Location Awareness which is somethign I also presumed was needed again except for the occassional test. The only service I know of with the word "discovery" in it is SSDP Discovery Service which enables discovery of UPnP devices which is something I typically keep disabled. I definitely did have NetBIOS over TCP/IP enabled as well as the had the TCP/IP NetBIOS Helper Service set to Automatic.

However it could very well have been a matter of not having a PC always powered up (something I really don't want to have to do) and simply not waiting long enough for the names to propogate through the system. I doubt I waited much longer than 5 minutes for the Workgroup entries to become availble before I reset the PC and tried something else.

If there really is a network Discovery service then perhaps that was my problem.

---

In the mean time I did implement Phoon's suggestion and that appears to work so far. I'm a bit hesitant to be overly certain and I want to see it work for at least a day or so before I declare victory but so far that has worked very well over a number of reboots of all the devices.

Basically the idea was to setup everything for dynamic IP addresses but simply reserve specific ones for the MAC addresses of the devices on the network.

I do have a few anomalies probably left over from some funky service tweaks or whatever those should be easy enough to fix.

I do have one thing I've noticed that hasn't appeared to make a difference but it is a bit concerning. When I last had everythig configured dynamically both the Primary DNS suffix and the Connection-specific DNS Suffix were empty now I have a Connection-specific DNS Suffix of 0.0.0.0 and I'm not sure where that's coming from nor am I sure that it's not a problem. If anyone has any knowledge about this then I'd appreciate it. The 0.0.0.0 also appears as the DNS Suffix Search List, not sure if this is something else or essentially a duplication of the connection specific DNS Suffix.

So the bottom line is at this point I do appear to at least have a workaroud however it would be good to know why it seems to be so difficult to have a consistent Workgroup when using fixed IP addresses and what one could do if you really needed fixed IP's.

I'm probably missing the boat here, but your router LAN configuration dialog should allow you to both use the router as a DHCP server and reserve specific IP addresses for specific devices - you can have static IP addresses designated for your networked PC's and networked (cabled or wireless) printers/scanners/etc.

I don't use a/the Computer Browser service so I'm in the dark there, but I'm not sure how that relates to assignment of LAN IP addresses by the router.

 

I didn't mean that there was a service called "network discovery" I meant certain services collectively used to perform "network discovery" such as:

 

DNS client service

FDResPub service (Function Discovery Resource Publication)

SSDPSrv and UPnPhost services

 

Generally speaking those 4 services make up "network discovery" on newer windows operating systems.  Older windows operating systems don't refer to it as "network discovery" but the end result using Netbios and workstation broker services is the same.

 

 

Yes it did and at this point that is my workaround. The issue was that the "manual" for the router didn't mention this at all and the menu involved with reserving IP addresses was not totally obvious. However when Phoon suggested this exact thing above then I poked around until I found it. But without someone's suggestion I had no way to simply intuit this functionality.

The other point in all this is why can't you have a Workgroup setup with fixed IP's and still have all members of the workgroup reliably visible to each other? Seems like someone would have stumbled over this issue before but perhaps networks with a bunch of fixed IP addresses like those at work are managed in a domain with dedicated domain servers. I'm just trying to network a pair of PC's and a Printer/scanner with a universal media card reader together on the cheap and had no previous experience with a router.

I'm not familiar with FDResPub it's probably Win 7 services. I'm still on XP.

Also from what I've been told UPnP is basically evil and should be disabled although at the moment I don't recall precisely why.

I generally try to avoid aotumatically starting services that therefore run at all times whether they're needed or not. A manual service that starts up when needed and shuts down when finished is a thing of bueaty to me too bad it's the rare service that actually operates like that. There are quite a few you can set to manual and they'll stay out of your way until needed and then they' trun on but they forget to turn themselves off afterwards and those are still far better than services that must be set to automatic.

Take the print spooler for example. I may actually need to print one out of every ten times I power up my PC. So why should I have my print spooler set to run automatically? When I need to print I just take the extra step to start the service. Most people would say why bother and if all we were talking about was a single service I would agree however if you use that attitude you'll soon fing yourself with 50+ processes always running with no idea of what half of them even do.

I've pretty much kept myself down to 15 processes on bootup including my security suite. His is with XP Pro 32 bit I'm sure Win 7 would require quite a few more than this but still if you don't care what operates on your PC then it will control you as opposed to the other way around.

My setup sounds just like yours, Mumble.  Still on XP here, too.  I have a desktop to which the router is cabled (cable modem -> router -> desktop) with a printer (ethernet) cabled to the router.  My other 2 laptops connect via wireless.  I'm a little anal about security, so I use MAC filtering, non-broadcast SSID & WPA2 encryption.  My printer is 'local' using a TCP/IP port.  My workgroup is the same on each PC (the one that shows up in the My Computer properties dialog).

Once I've entered all the devices & assigned reserved IP addresses in the router dialog, I enable drive sharing on the desktop & notebooks for the drives I want the other PC's on the network to 'see' and then map to them on each respectively.  The printer is installed on each PC using the same TCP/IP port assignment.  This all seems to work.  Don't have a stand-alone card reader, but most I've seen are USB cabled to a PC; a drive letter is auto-assigned to each individual slot of the reader and those can be shared like any HD or USB backup drive, etc.

Maybe that helps?

Not sure about this. Is this a router function?

I had this as an option as well I could have connect all three over USB for Printer/Scanner access. I also could have used the USB as access to the Printer's card reader but I would have had to install basically what amounts to a USB network on each PC simply to get to the card reader.

All of this seemed like a bit too much in added cabling and software when I clearly needed at least the PC's to have ethernet connections to the router seemed like a natural to connect the printers ethernet directly to the router as well.

Do you actually map them as a network drive that's reconnected on logon? If so this is actually an easier way that doesn't really need everything that you need to be able to see the workgroup reliably under all conditions.

 

---

 

My issues all probably stem from being very particular about what services run on my PC as well as what software runs and what ancillary processes/software is required by any new piece of software that I might otherwise be interested in.

I'm not gullible enough to believe that uninstalling some software that does something I don't want it to do actually really cleans up after itself and though I do have a pretty decent and aggressive registry cleaner I prefer to avoid the situation entirely wherever possible. I also don't trust restore points or any backup service that requires the OS be "live" during the backup.

The only kind of backup I trust is the Norton Ghost 2003 that I have that boots from the floppy and takes a boot drive image while the OS is inert. That what I do before embarking on any kind of software install or system modification like this. That way if something bad happens I can always get back to the *identical* state prior to the modifications without any potential time bombs left in the registry or an otherwise incomplete uninstall.

Anyway at this point, I think I'm straightened out but it was a real pain in the ass.

The bottom line was that I had no issue with either PC accessing the internet nor any problem printing to or scanning from the Printer under any of the myriad (which literally means ten thousand) combinations of settings and services that I checked.

The only issue was involved in being able to access shared directories between the two PC's as well as be able to access the universal media card reader in the Printer.

One thing I definitely found out was that if you setup your network to use static IP's there was no way,at least that I could find, that allowed you to reliably access all three members of the Workgroup.

The reason I needed to use static IP's as opposed to the more standard dynamic assignment through a DHCP server (usually provide by the router) was because when I used this method and I tried to scan the network for a network attached printer for whatever reason the Printer setup wizard could not find it so I had to resort to setting up a hard coded IP printer port on each of the PC's. This was fine until the lease on the IP's ran out and for whatever reason the Printer was not assigned the same IP. At first I simply configured another IP printer port and changed the printer to that which worked fine for the printer but not the scanner. To get the scanner to work I had to uninstall the Printer/Scanner and then reinstall to the new printer port. This got to be a major pain in the ass even with the lease time set to its maximum of 48 hours.

So then I tried static IP's but like I mentioned above there was no way that I could find that allowed you to reliably access all three members of the Workgroup. I could see some devices some of the time under various but like I said nothing was reliable from one boot to the next.

Si then I discovered that I could continue to use the DHCP seerver in the modem and have it allocate IP's but also I could reserve an IP address for the MAC address of a specific device in the system. This essentially made the lease's permanent although in the status the lease was said to be valid only until 2038 but since I'll be dead by then I figure that's close enough.

Even with these changes I wasn't quite out of the woods yet. I'm a notorious service tweak hound and previously I had my default configuration down to 15 services including my full security suite running, 16 services if I had a browser open and that was a perfectly stable situation for years prior to switching over to DSL service.

Anyway I found a few things that I didn't previously know about services. For one turning a service on after boot is entirely different than having it start automatically. Basically starting up a service after boot may in fact cause some function to work but if you start the service automatically it might not work. You also get the converse where starting up a service after boot doesn't cause something to work but starting it automatically does. In order to avoid having to do something else before you can do what you want I prefer to have everything I want work right from bootup.

Another thing is that although the descriptions do give some general guidance it's not totally clear whether you need something that from the description you would think has to be required to make something work nor is it certain that you don't need something simply because it's description has nothing to do with the thing in question nor is it necessarily a required underpinning of something that you do need.

Basically if you're going to tweak services you need to change one thing at a time and then reboot before testing a feature then once you have that feature working you can again turn off one service at a time then reboot to see if that feature still works.

Also note that not every service that you turn on actually causes a new service to appear under Taskman. For example a number of services turn on svchost.exe under the SYSTEM user. Others create an instantiation of svchost under NETWORK SERVICE and others create a svchost under LOCAL SYSTEM. Often starting another service that shares the same service and username doesn't actually add a new service it just makes the previous service just a bit larger, but other times it might just and another instantiation of the service under the same username.

The bottom line of the services that I need to explicitly set to Automatic, under XP Home are the following. In XP Pro I did not apparently need DNS Client but since I needed it on the other PC I set it to Automatic on the XP Pro system even though it cost me an "extra" instantiation of svchost.exe running under the NETWORK SERVICE user.

Computer Browser
DNS Client
Network Location Awareness
WebClient

The converse of this is that if you don't have a local Workgroup you can turn these services off (and probably a few more that I'm not aware of).

All of this just to get a local workgroup running.

One other discovery I made related to this was that if you open an Explorer window and go to View tab the very first option under Files and Folders is to "Automatically search for network folder and printers" if you check this box you actually get an item directly under My Network Places > Entire Network without having to actually dive down into Microsoft Windows network > Workgroup and in many cases provide access to all the shared folders in the network even if all of the network devices haven't been "discovered" yet for whatever the reason. While this is a coll feature I didn't want to count entirely on it hence the instance that all network devices be visible for the others through the normal Workgroup path.

All of the above seem more related to simply being able to find the devices and directories under them because in pretty much all cases the path was "there" already even though the PC might not know it. For example even without the above services you could simply open an Explorer window and map a network drive simply by typing the known path in and browse to the directory. You don't even have to complete the mapping but now suddenly that device would appear in the Workgroup and be accessible via that path.

All in all two total extra services to have this reliable under all conditions without having to do anything funky to get it to work was worth it to me. To most people that probably have 30 to 40 default services running from the start probably would never notice any of this, but then they'd be relegated to running a whole bunch of crap that they don't need and have no knowledge of and that's something I'm not willing to do even though it causes me the occasional problem like this. But then that's another story.

 

Incorrect.  I only run static IP's and have never had an issue whatsoever, of course I also don't agressively service-manage (ala internet-joe's like Black Viper etc.) so "network discovery" has never been an issue for any of my systems.

Choosing STATIC or DYNAMIC IP's has absolutely nothing to do with the success or failure rate of "network discovery".........if anything it had to do with your service-managing just so you know. 

The other thing with services.  You said you like to limit things.  No problem there except for the fact that people like Black Viper etc. who talk about which services are needed and not do not give the full list of service dependencies (they can't because MS hasn't in many cases provided them either) so people end up disabling a service that handles A (which they've determined they don't need) but the service also performs or is required for reliable B operation (which the person doesn't know about because MS's service description or dependencies list is incomplete).

My suggestion as always is to only disable services after you've educated yourself enough about the individual service and any/all potential depedencies in any myriad of situations (which can of course become rather time-consuming). 

When I ran XP (I didn't have a need for a workgroup back then) my system booted in 21 seconds to a complete desktop and at idle had exactly 12 services running in the taskmgr.  Now that I'm running Windows 7 my system still boots to a clean desktop in under 30 seconds but I have 32 services running at idle.  Of course I also have 8GBs of RAM and ooodles more processing power since that XP machine so I'm not as agressive with service-management these days.  At the end of the day what I'm saying is:  It's easy to kill services, just make sure you're not killing the usability of your system at the same time.

 

the Monk