Windows/IIS troubles

[a_guy]

Hello fellow ServerBeach subscribers :

First and foremost, being an AIX/UNIX guy, I feel totally lost on Windows. Seems as though none of the command-line tools are available, and the init scripts aren't in place, it's driving me bonkers.

I've inherited a Windows server at work, and for the most part, things seem to be behaving like I'd expect (after reading the manuals, which aren't in the man pages for some reason), except a small problem. I've recently put up ~175+ domains, and 15 of those don't want to resolve at the IIS/Server level. I know that DNS is pointing to the right place and gives me the correct IP when I try to ping, The file system permissions are permissive to the degree that IIS can read the files, the host headers are in agreement to the FQDN's, the common names agree with the host headers in IIS (I don't know if this is a requirement, but it's a standard I like to use), the Directories are all named by the common name less the .com extension (works for the other 160).

For some reason, the server itself doesn't seem to recognize that it is responsible for those names, and it times out even when I ping it from itself using it's IP address, or one of the FQDN's that are misbehaving.

Configuration:
OS - Win 2000 Server
Web - IIS 5?
DNS - ns1/ns2.geodns.net
In ServerBeach DNS-Tool everything looks right
Registrar - GoDaddy - DNS points to geodns
Hosts File - localhost is the only thing registered

Firefox gives me a timeout error, so does ping. It's both local and server-side

What in this universe could I have overlooked?

Any help would be much appreciated.

P.S. I know it's been said before, but if there was a way to use the DNS-Tool in bulk, that would be highly convenient.

[chavez]

Hello fellow ServerBeach subscribers :
Welcome!
I've inherited a Windows server at work,
Sorry to hear that.. ;)
For some reason, the server itself doesn't seem to recognize that it is responsible for those names, and it times out even when I ping it from itself using it's IP address, or one of the FQDN's that are misbehaving.
This is the part that bugs me. If you can't ping the IP address while you're on the server, you might want to remove and reconfigure the IP. Also, I seem to remember the firewall on Win2K causing problems with multiple IP addresses on one interface. Take the firewall out of the picture while your testing this.
P.S. I know it's been said before, but if there was a way to use the DNS-Tool in bulk, that would be highly convenient.
I agree! :) Thanks for the suggestion.

[a_guy]

This is the part that bugs me. If you can't ping the IP address while you're on the server, you might want to remove and reconfigure the IP. Also, I seem to remember the firewall on Win2K causing problems with multiple IP addresses on one interface. Take the firewall out of the picture while your testing this.

Thanks for the quick reply Chavez, and thanks for the welcome. I'm not sure how to re-bind the IP address on a Windows machine, ifconfig doesn't want to run (so I'll have to assume it's not present). I'm not sure it is the IP address, since when I go to ping the well-behaving FQDN's, they respond with vigor, both locally and from the server.

I also don't think it has anything to do with application pooling, which I've read may have caused this issue from other places on the board. I've moved the names from the "Default Application" pool to another pool called "Root" and still no dice. The only thing that these names have in common is that they were forwarded at the registrar until I got my hands on them. This is also true of some but not all of the other domains that are behaving well. I'm also reluctant to blame the firewall, since there are ~160 names that do behave correctly, and since that's the case, I would imagine that a firewall would behave similarly on Windows and either allow/deny access to a resource. Being an old UNIX hand I could be wrong about the behaviours exhibited on Windows, but I strongly doubt that firewalls would behave erratically and still be called firewalls.

I'm willing to re-bind the IP address if anyone could give me a head's up on how that's done on the Windows side. While I don't think it will solve the problem, since it isn't solved, it can't hurt anything. Aside from that, does anyone have any other ideas on why 15 of 175 virtual hosts would misbehave? I've been bashing my head for the last 2 days on this, but it feels like I'm talking to a brick wall when I talk to Windows :bang:

[chavez]

ifconfig is a Unix command. Windows uses ipconfig, but you may as well use the GUI. At any rate, if this is a server with a single IP address, you can probably disregard the suggestion to reconfigure the IP address.

It doesn't make sense that the correct IP is displayed when you ping the non-working domain, but the server doesn't respond. Yet when you ping a working domain, the server does respond. Do both domains point to the same IP address?

Based simply on what you describe, you're not experiencing any of the usual problems. Incorrect host headers usually result in a Bad Request (Invalid Hostname) error and application pool problems will give a Service Unavailable error. There will also be errors in the Event Viewer about App Pool problems.

What happens when you browse to the site? If it times out, you may have code or database issues. Try adding a simple HTML file to the site's root folder and browse the direct URL for this file. Also try browsing the site directly from IIS (right-click a site file and select "browse").

That's all the random ideas I can think of for now. :D Good luck!

[DXD]

Thanks for the quick reply Chavez, and thanks for the welcome. I'm not sure how to re-bind the IP address on a Windows machine, ifconfig doesn't want to run (so I'll have to assume it's not present). I'm not sure it is the IP address, since when I go to ping the well-behaving FQDN's, they respond with vigor, both locally and from the server.

I also don't think it has anything to do with application pooling, which I've read may have caused this issue from other places on the board. I've moved the names from the "Default Application" pool to another pool called "Root" and still no dice. The only thing that these names have in common is that they were forwarded at the registrar until I got my hands on them. This is also true of some but not all of the other domains that are behaving well. I'm also reluctant to blame the firewall, since there are ~160 names that do behave correctly, and since that's the case, I would imagine that a firewall would behave similarly on Windows and either allow/deny access to a resource. Being an old UNIX hand I could be wrong about the behaviours exhibited on Windows, but I strongly doubt that firewalls would behave erratically and still be called firewalls.

I'm willing to re-bind the IP address if anyone could give me a head's up on how that's done on the Windows side. While I don't think it will solve the problem, since it isn't solved, it can't hurt anything. Aside from that, does anyone have any other ideas on why 15 of 175 virtual hosts would misbehave? I've been bashing my head for the last 2 days on this, but it feels like I'm talking to a brick wall when I talk to Windows :bang:

If you want to PM me the 1 or 2 of the working domains names and their IP and the names of the none working domain names and what there ip's are suppose to be I'll diagnose and figure out your problem for you.

[a_guy]

ifconfig is a Unix command. Windows uses ipconfig, but you may as well use the GUI. At any rate, if this is a server with a single IP address, you can probably disregard the suggestion to reconfigure the IP address.

It doesn't make sense that the correct IP is displayed when you ping the non-working domain, but the server doesn't respond. Yet when you ping a working domain, the server does respond. Do both domains point to the same IP address?

Based simply on what you describe, you're not experiencing any of the usual problems. Incorrect host headers usually result in a Bad Request (Invalid Hostname) error and application pool problems will give a Service Unavailable error. There will also be errors in the Event Viewer about App Pool problems.

What happens when you browse to the site? If it times out, you may have code or database issues. Try adding a simple HTML file to the site's root folder and browse the direct URL for this file. Also try browsing the site directly from IIS (right-click a site file and select "browse").

That's all the random ideas I can think of for now. :D Good luck!

Chavez, The problem with these troubleshooting steps is that all files are plain vanilla HTML files. I don't get a Bad Request (Invalid Hostname) or a Service Unavailable error, it just times out. I'm also running on just 1 IP address, since I don't need SSL for these domains, they're just alternate entry product lines for the company. Here's the Firefox message when I try to open the root (default page is index.html):


The connection has timed out


The server at www.intentionallyleftblank.com is taking too long to respond.


* The site could be temporarily unavailable or too busy. Try again in a few
moments.

* If you are unable to load any pages, check your computer's network
connection.

* If your computer or network is protected by a firewall or proxy, make sure
that Firefox is permitted to access the Web.


On the same IP address, the site http://www.protectedname.com displays it's index.html without issue. It would be great if this were simply a standard issue, as I have been troubleshooting web applications since before most people knew there was a "web" (Go 1200 baud couplers! Page load average a screaming 7 minutes flat!). This is the first time that I've had anything even remotely like this happen, but so happens when they make browsers anyone can use, and get rid of the useful troubleshooting information. I decided to post for help after I'd exhausted all other resources that I'm aware of, and after checking everything end-to-end on this from the vhost config (Host Headers), to DNS, to other connections to the same service.

I'm sure I've goofed on a setting, but since I am un-familiar with the daemon and the platform, it gets frustrating after reading the manuals, double-checking that the settings are identical for everything, and still no warm-fuzzy that the scenario has changed.

How would I go about re-binding the TCP/IP stack with a GUI tool, or is that even something left to consider considering that 160 work fine?
Is there any other information I can offer?
This is a highly unusual happening from my experience, and that fact that you guys are doing such a great job in trying to assist shows me a great deal from ServerBeach. I may even move some of my personal serving needs over here if I can expect such a helpful resource as this forum.

Thanks for the heads-up on the windows ipconfig utility, it doesn't seem too much like ifconfig, but it may be useful in the event I need to get the TCP/IP stack rebuilt.


If you want to PM me the 1 or 2 of the working domains names and their IP and the names of the none working domain names and what there ip's are suppose to be I'll diagnose and figure out your problem for you.

DXD, I'd love to PM you, and I think that a solution that benefits all is much more useful than a solution that benefits a single user. I'll send you the names, as long as I get to re-post your answer to the board with the innocent domain names protected for their safety (I don't want to see any domains injured in the solving of this problem)

[DXD]

Chavez, The problem with these troubleshooting steps is that all files are plain vanilla HTML files. I don't get a Bad Request (Invalid Hostname) or a Service Unavailable error, it just times out. I'm also running on just 1 IP address, since I don't need SSL for these domains, they're just alternate entry product lines for the company. Here's the Firefox message when I try to open the root (default page is index.html):


The connection has timed out


The server at www.intentionallyleftblank.com is taking too long to respond.


* The site could be temporarily unavailable or too busy. Try again in a few
moments.

* If you are unable to load any pages, check your computer's network
connection.

* If your computer or network is protected by a firewall or proxy, make sure
that Firefox is permitted to access the Web.


On the same IP address, the site http://www.protectedname.com displays it's index.html without issue. It would be great if this were simply a standard issue, as I have been troubleshooting web applications since before most people knew there was a "web" (Go 1200 baud couplers! Page load average a screaming 7 minutes flat!). This is the first time that I've had anything even remotely like this happen, but so happens when they make browsers anyone can use, and get rid of the useful troubleshooting information. I decided to post for help after I'd exhausted all other resources that I'm aware of, and after checking everything end-to-end on this from the vhost config (Host Headers), to DNS, to other connections to the same service.

I'm sure I've goofed on a setting, but since I am un-familiar with the daemon and the platform, it gets frustrating after reading the manuals, double-checking that the settings are identical for everything, and still no warm-fuzzy that the scenario has changed.

How would I go about re-binding the TCP/IP stack with a GUI tool, or is that even something left to consider considering that 160 work fine?
Is there any other information I can offer?
This is a highly unusual happening from my experience, and that fact that you guys are doing such a great job in trying to assist shows me a great deal from ServerBeach. I may even move some of my personal serving needs over here if I can expect such a helpful resource as this forum.

Thanks for the heads-up on the windows ipconfig utility, it doesn't seem too much like ifconfig, but it may be useful in the event I need to get the TCP/IP stack rebuilt.



DXD, I'd love to PM you, and I think that a solution that benefits all is much more useful than a solution that benefits a single user. I'll send you the names, as long as I get to re-post your answer to the board with the innocent domain names protected for their safety (I don't want to see any domains injured in the solving of this problem)

That's fine, I just didn't want you to have to post all that private information here.

[a_guy]

That's fine, I just didn't want you to have to post all that private information here.

Thanks mate, You're a great sport for letting my woes help others and I commend you for that. You've gained considerable respect in my book, since most people are very much "This is MY solution, and while it could help others, it's MY property". Thank you for exceeding my expectations of this community, and I can honestly say it feels like a community effort with assistance from helpful selfless people that have talent in this arena. Reminds me of SourceForge and that has to be one of my favourite communities with this one becoming a new favourite.

[a_guy]

These domain names are pointing to the wrong IP address (65.34.X.X). This can be changed in the DNS Tool.

name1.com
name2.com
name3.com
name4.com
name5.com
name6.COM
name7.COM
name8.COM
name9.COM
name10.COM
name11.COM


These sites give the error:
"Directory Listing Denied"
This means you need to allow directory listing in IIS, or create a default page, such as index.htm .
name12.com
name13.com
name14.com
name15.com

I haven't looked at the last 4 for a few days, so if they're giving a Directory Listing Denied, I can get that solved straight away. I feel like an idiot for not noticing the 65 in the place of the 64, all of the other servers I manage are on the 65.X.X.X block, so I must have just overlooked it since it looked good compared to all of the other servers I run to ping. I knew it had to be something simple that I was overlooking, so that second set of eyes really helped a LOT.
I'll post this information to the public boards, so the next time someone is scratching their heads and bashing them against the wall, they'll know that ANYONE can overlook the most basic configuration setting. Thanks mate, you've made my week. Looks like I'll be moving my personal business to ServerBeach as well, So long as there are no qualms about a Gentoo or Slackware install (I refuse to use Windows, and I like my kernel how I like my kernel - LOL).
:applaud: :thanks: