Symptom of running out of subnet?
-
Does the A record by chance have more than one IP that could be causing issue?
Do you actually have an A record for it or are you relying on NetBIOS lookup for the Hostname instead of DNS? -
@JaredBusch said:
Cycling DNS has nothing to do with running out of IP addresses in your current subnet.
Assuming that your DNS server properly knows the hostname and IP of the application server, you should never have this kind of problem.
Have you tested that DNS is actually broke when a client cannot connect? I mean is a ping or nslookup returning bad information from a client that cannot use the application?
I had a typo. I meant to say "reside on a different subnet ip address"
I will test ping next time it happens.Is there a way to start DNS fresh? I inherited it from my predecessor, and we used to do a lot more with IT than just DC.
-
@LAH3385 said:
@JaredBusch said:
Cycling DNS has nothing to do with running out of IP addresses in your current subnet.
Assuming that your DNS server properly knows the hostname and IP of the application server, you should never have this kind of problem.
Have you tested that DNS is actually broke when a client cannot connect? I mean is a ping or nslookup returning bad information from a client that cannot use the application?
I had a typo. I meant to say "reside on a different subnet ip address"
I will test ping next time it happens.Is there a way to start DNS fresh? I inherited it from my predecessor, and we used to do a lot more with IT than just DC.
If you are on a different subnet, then are you certain that both pieces have a scope large enough to talk to each other? Did the server get updated to the new subnet mask as well as the endpoint? Is there a DB server involved also? did it get updated too?
-
@JaredBusch
I am 90% certain it is not enough. Both subnet mask are still /24. I have roughly 170 devices on subnet1 and 90 devices on subnet2. -
@LAH3385 said:
@JaredBusch
I am 90% certain it is not enough. Both subnet mask are still /24. I have roughly 170 devices on subnet1 and 90 devices on subnet2.Is there a router between the two?
-
@Jason said:
@LAH3385 said:
@JaredBusch
I am 90% certain it is not enough. Both subnet mask are still /24. I have roughly 170 devices on subnet1 and 90 devices on subnet2.Is there a router between the two?
Need more coffee... Yeah, you are already on 2 different subnets? Then are you sure you are routing correctly? Also, you will have to be relying on DNS and not NetBIOS if this is really true.
-
@Jason
Not that I am aware of. They are connected by a switch.@JaredBusch said in Symptom of running out of subnet?:
@Jason said:
@LAH3385 said:
@JaredBusch
I am 90% certain it is not enough. Both subnet mask are still /24. I have roughly 170 devices on subnet1 and 90 devices on subnet2.Is there a router between the two?
Need more coffee... Yeah, you are already on 2 different subnets? Then are you sure you are routing correctly? Also, you will have to be relying on DNS and not NetBIOS if this is really true.
All the devices on subnet2 are statics. They are the phone server and the ip phones. The application in question is to keep log of calls (inbound/outbound/missed/etc)
I did configure a gateways to our Firewall (watchguard) switch. Each subnet has its own gateway. -
-
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
I need to expand subnetmask pretty soon. Hoping to do it on long weekend. Give me enough time to fix any problem that may arise.
-
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
Look at the PC that is broke and the application server. What are their IP settings?
Example:
The server is 192.168.1.2/24 with a gateway of 192.168.1.1
The desktop is 192.168.2.2/24 with a gateway of 192.168.2.1
Those are two devices on different subnets.In order to ping from 192.168.2.2 to 192.168.1.2, the system will route it through the gateway because the destination is on another subnet.
-
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
Look at the PC that is broke and the application server. What are their IP settings?
Example:
The server is 192.168.1.2/24 with a gateway of 192.168.1.1
The desktop is 192.168.2.2/24 with a gateway of 192.168.2.1
Those are two devices on different subnets.In order to ping from 192.168.2.2 to 192.168.1.2, the system will route it through the gateway because the destination is on another subnet.
In that case I believe our Firewall also act as router. Because we are able to ping ip phones
-
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
Look at the PC that is broke and the application server. What are their IP settings?
Example:
The server is 192.168.1.2/24 with a gateway of 192.168.1.1
The desktop is 192.168.2.2/24 with a gateway of 192.168.2.1
Those are two devices on different subnets.In order to ping from 192.168.2.2 to 192.168.1.2, the system will route it through the gateway because the destination is on another subnet.
In that case I believe our Firewall also act as router. Because we are able to ping ip phones
Your "firewall" is a router with additional firewall capabilities.
-
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
Look at the PC that is broke and the application server. What are their IP settings?
Example:
The server is 192.168.1.2/24 with a gateway of 192.168.1.1
The desktop is 192.168.2.2/24 with a gateway of 192.168.2.1
Those are two devices on different subnets.In order to ping from 192.168.2.2 to 192.168.1.2, the system will route it through the gateway because the destination is on another subnet.
In that case I believe our Firewall also act as router. Because we are able to ping ip phones
Your "firewall" is a router with additional firewall capabilities.
That make sense when you put it that way.
So do you think subneting may fix the dns cycle issue? What does a good/clean DNS looks like? -
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.@JaredBusch said in Symptom of running out of subnet?:
@Jason said:
@LAH3385 said:
@JaredBusch
I am 90% certain it is not enough. Both subnet mask are still /24. I have roughly 170 devices on subnet1 and 90 devices on subnet2.Is there a router between the two?
Need more coffee... Yeah, you are already on 2 different subnets? Then are you sure you are routing correctly? Also, you will have to be relying on DNS and not NetBIOS if this is really true.
All the devices on subnet2 are statics. They are the phone server and the ip phones. The application in question is to keep log of calls (inbound/outbound/missed/etc)
I did configure a gateways to our Firewall (watchguard) switch. Each subnet has its own gateway.Yes, subnets would need a gateway to be networked to anything. How do they talk to each other, though?
-
@LAH3385 said:
In that case I believe our Firewall also act as router. Because we are able to ping ip phones
For all intents and purposes, firewalls and routers are the same thing. It's been decades since anyone made a router without firewall capabilities. The terms are really interchangeable especially in the SMB.
-
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
Look at the PC that is broke and the application server. What are their IP settings?
Example:
The server is 192.168.1.2/24 with a gateway of 192.168.1.1
The desktop is 192.168.2.2/24 with a gateway of 192.168.2.1
Those are two devices on different subnets.In order to ping from 192.168.2.2 to 192.168.1.2, the system will route it through the gateway because the destination is on another subnet.
In that case I believe our Firewall also act as router. Because we are able to ping ip phones
Your "firewall" is a router with additional firewall capabilities.
That make sense when you put it that way.
So do you think subneting may fix the dns cycle issue? What does a good/clean DNS looks like?No, subnetting and DNS should not be related. If tehre is a problem like this, likely it will get carried through.
-
@scottalanmiller said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
It is that kind of moment that somehow it works momentary, but you cannot sustain it, but it works, but you don't know how is it working, but it works...
Look at the PC that is broke and the application server. What are their IP settings?
Example:
The server is 192.168.1.2/24 with a gateway of 192.168.1.1
The desktop is 192.168.2.2/24 with a gateway of 192.168.2.1
Those are two devices on different subnets.In order to ping from 192.168.2.2 to 192.168.1.2, the system will route it through the gateway because the destination is on another subnet.
In that case I believe our Firewall also act as router. Because we are able to ping ip phones
Your "firewall" is a router with additional firewall capabilities.
That make sense when you put it that way.
So do you think subneting may fix the dns cycle issue? What does a good/clean DNS looks like?No, subnetting and DNS should not be related. If tehre is a problem like this, likely it will get carried through.
Our DNS was inherited from an old company (same CEO). I do not know what has been done, and I don't usually mess with DNS. All I know is DNS works with DHCP in some way. Is there something I might be missing?
-
@LAH3385 said:
All I know is DNS works with DHCP in some way. Is there something I might be missing?
Only in the fact that DHCP hands out DNS, and DHCP can register in DNS if you are using windows DHCP/DNS.
No more related than the fact that DHCP can handout TFTP or SIP config but it is not tied into them.
-
@JaredBusch said:
@LAH3385 said:
@Jason
Not that I am aware of. They are connected by a switch.You cannot access TCP/IP across two different subnets without something routing traffic.
While this is true - a Layer 3 switch can route between subnets - that's what I do for my VLANs. I don't have a router handle this, I have a my layer 3 switch handle it.
-
From the computer having the problem, run
tracert xxx.xxx.xxx.xxx
Replace the x's with the IP of the server.
Do that same thing from the server to the PC's IP address.
Post the results if you can.
Also post the results from ipconfig /all from both.