New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Proxmox 4.x/SoYouStart
I am having issues configuring the networking for a VM running Win 2008. I checked several guides, but I must be missing something.
Here is what I have done.
1. Created a virtual mac address in the SYS Panel for the failover IP I am using
2. Created the network device for the VM in Proxmox including the mac address using Virtio
3. Entered the IP, Subnet Mask, Gateway, and OVH DNS in the Windows network settings.
Does anyone see what I'm missing?
Comments
Have you make bridge interface in the host?
Read this carefully
http://help.ovh.co.uk/Proxmox
Yes, that is the guide I used.
IP : 1.2.3.0
Subnet Mask: 255.255.255.255
Gateway: 1.2.3.254
DNS: 213.186.33.99
Do you have another vm and its networking works?
I tried and still no luck. Not sure what I have missed.
Maybe this'll help, maybe not, but here's some notes I saved when setting up a Windows Server 2016 VM with a failover IP on my SoYouStart box:
I'm not sure why the registry values shit was necessary but I was being provided a local IP (169.254.x.x) until I added them. Also another thing that tripped me up is that you need to assign the main linux bridge in proxmox (the one with your host IP) to the VM, not the one with your failover IP.
Gateway have to be the host IP with last octet at 254.
https://docs.ovh.com/au/en/cloud/dedicated/network-bridging/
In Windows, you might need to set the netmask to 255.255.255.0, apply the setting and then set it to 255.255.255.255 (which is important or your server is going to send broadcast on the network and you'll get a mail from our monitoring )
Should work out of the box if you are using the correct MAC and driver for your Windows guest
Funny you should ask this.... I just went a few rounds with this today with a Xenserver/Untangle combo trying to get it working. Done it many times before, but often times different providers require different things.
I was driving myself crazy thinking typed in something wrong(IP, gateway, subnet, MAC), there was a problem with my firewall settings, etc....
MaikoB is 100% correct.
My issue was I was reading through their documents too fast and missed the part saying to use the main IP's gateway for the failover IP's gateway.
In other words:
main IP = 1.2.3.4
gateway = 1.2.3.254
failover IP = 9.8.7.6
gateway = 1.2.3.254
Plugged that in and bingo.... instant pings....
They talk about it a bit here in this doc. You can extrapolate most of the info from it. I was just reading to fast to catch that one very important part.
http://help.ovh.ie/BridgeClient
Hope that helps!
Were you able to get it to work?
As said above:
Netmask: 255.255.255.0 might work.
Or once I clicked diagnose and windows self-healed the issue.
yes, more details on what you did and how you actual configuration looks like are missing ;-)
please elaborate about you bridge setup, maybe simpliest post the whole /etc/network/interfaces and ouput of ifconfig (both from the node of course)
for the gateway: as other pointed out you should use the gateway for the main IP of the the host which will be that IP ending in .254
also the netmask most probably should not be 255.255.255.0 but 255.255.255.255 to have the guest use static routing (can't exactly tell how this is implemented in win 2008 though)
that's just an urban myth totally not to believed in. just use google dns 8.8.8.8 or any other existing one and you're good. after all definetly not the point of failure.
this is a good question, so an answer to that would be nice - just to narrow down if this problem is related to your general setup or specific to that single windows VM.
I would not really recommend that for a server.
Using our DNS server is recommended because of it's low latency in our Network : 213.186.33.99
I've seen people using strange DNS server and then wndering why this or this action take so much time... (like sending mail which require multiple DNS request) I even recommend to set your own DNS Servers and use it in your guest
while your reason behind your recommendation totally makes sense, I've never ran into problems using the google dns so far. I use OHV dns as second entry though...
here I was more about the story that using OVH dns is a MUST - which clearly isn't true ;-)
Thanks everyone! I was not using the gateway for the main ip.
@MaikoB
What would cause pages in Chrome to load very slowly? I'm using the OVH DNS and pings to the sites are normal.
Hmmm is it only chrome or all browsers?
That's what I did, set up a DNS resolver, forward all DNS requests to 213.186.33.99 and then point guest OSs to the internal IP of the DNS server you're running.
No issues at all, plus DNS names are already cached locally on the DNS server itself and anything else is forwarded to OVH's server.
Even though it's lower latency, I found it took WAYYY longer to do standard lookups vs OpenDNS or Google DNS. Those servers must be getting hammered, lol. Even though Google DNS is like +20ms away, it still returns results way faster than OVH DNS from my testing.
Really ?
$ dig lowendtalk.com @8.8.8.8 | grep "Query"
;; Query time: 129 msec
$ dig lowendtalk.com @8.8.8.8 | grep "Query"
;; Query time: 135 msec
$ dig lowendtalk.com @8.8.8.8 | grep "Query"
;; Query time: 129 msec
$ dig lowendtalk.com @213.186.33.99 | grep "Query"
;; Query time: 0 msec
$ dig lowendtalk.com @213.186.33.99 | grep "Query"
;; Query time: 0 msec
$ dig lowendtalk.com @213.186.33.99 | grep "Query"
;; Query time: 0 msec
$ dig lowendtalk.com @213.186.33.99 | grep "Query"
;; Query time: 0 msec
But yes, I agree, it depend on the cache. Google (and other like OpenDNS) get lot of request. The probability that you request something which is not in cache is low.
We get less request so you might get longer query time because the information might not be in cache already. But as soon as it's in cache, it's faaaast
Perhaps I just had some bad luck at BHS previously, was quite a bit slower when testing amongst a variety of queries. The thing that tipped me off to it was the rDNS lookups OpenSSH does for the connecting IP, was always +3-4s to connect. Swapped to Google DNS and it was always near instant shrug
The bigger list of cached results makes a difference with the weird lookups I apparently do, haha.
remember seeing that too :-)
Yeah it just depends on a lot of different factors
I am having a new issue after getting the network config resolved. I have a Win vm and a Debian 9 vm created in Proxmox. There is a web app running on the Debian vm that I can't access from the Win VM using any browser. I can access the app from my pc. I can access other sites from the Win vm. Any ideas?
Ping the Debian VM from the Windows VM, do you get a reply?
are the IPs of both VMs in the same subnet? what subnet did you enter in your network config on both ends?
The ping times out.
They are both in the same subnet.
1.2.3.0
1.2.3.1
@xenos: what netmask did you enter in debian and windows? 255.255.255.0 or 255.255.255.255 ? my guess would be that you have the former somewhere and therefore the vm is looking for a neighbour but can't reach it, because it would need to go through the gateway first.
The Win vm has 255.255.255.255 and I just noticed the Debian vm has 255.255.255.252. I changed it to 255.255.255.255 in /etc/network/interfaces but it reverts back to 252 after I reboot.
@xenos that's how my interfaces file usually looks like under debian:
where A.B.C.D is the IP assigned that VM and C.D.E.254 is the main gateway of the hostnode (same as discussed above for the windows settings)
you might try to run a traceroute from both ends, to see what it gives as an output...
@Falzo Mine is...
Traceroute to Win vm