Jeroen Landheer's questions

This question is more of a math question than a server question, but it is strongly server related.

If I have a server that I would be able to guarantee 95% uptime and I would put that server in a cluster of 2, how much would the uptime be then? Now, let's say I do the same, but I make it a cluster of 3?

Let's not consider things like single point of failure, but purely focus on the math here. One of the things that makes this a bit complicated is that if for example I have 2 servers, the chance that they are both off is 2^2, so that's 1/4th; or for 3 that's 2^3, so 1/8. Considering I have an 5% downtime for each of these servers, would the total average be then that 1/8th of that 5%?

How would you calculate something like this?

We're setting up a software firewall, and one of the issues we're facing is that eth0 has a dynamic IP address.

In /etc/network/interfaces we've got the following:

# The loopback network interface
auto lo
iface lo inet loopback

# internet-facing interface
auto eth0
iface eth0 inet dhcp
    dns-nameservers 192.168.5.8 192.168.5.5
    dns-search -------.local

# The primary network interface
auto eth1
iface eth1 inet static
    address 192.168.5.252
    netmask 255.255.255.0
    network 192.168.5.0
    broadcast 192.168.5.255
    dns-nameservers 192.168.5.8 192.168.5.5
    dns-search -------.local

# This is an autoconfigured IPv6 interface
iface eth1 inet6 auto

The problem is that resolv.conf ends up with the DNS IP's of our ISP...

Update As stated above, resolv.conf gets the DNS IP addresses of our ISP that are part of the DHCP Package. The idea is that we get just an IP address, but basically only use our internal DNS servers. So the DNS parameters of DHCP should be ignored. Any ideas how to accomplish this?

One of our clients asked us to configure Exchange 2010 to retain outgoing mail for a certain amount of time (independant of Outlook settings.) The idea is that an administrator has about 10 minutes to take a message out of a queue before it is sent out the organization.

I know this can be configured in Outlook, but this is not a valiable solution for us. I'm also aware that this causes queues to fill up, this is part of the consideration.

Is there a way in Exchange 2010 to configure this?

I'm testing a IPv6 setup using Hyper-V on 2008 R2.

I've setup one VM as a DC, with 2008 R2 Standard Edition, AD, DNS and DHCP. I've got one VM as Windows 7 RTM x64, as a domain member.

IPv4 connectivity works fine, no problems here. IPv6 on the other hand is acting strange. DHCPv6 works without a problem, the client gets an IPv6 address assigned and the DNS Server address is delivered as well. But once I ping using Ping -6 I get "PING: transmit failed. General failure". If I ping from the DC to the workstation I get a "Request timed out".

Since this is a test network I've disabled the firewalls, but this does not change the situation. I use fd48:5884:e6ae:e8e9::/64 as a prefix, fd48:5884:e6ae:e8e9::1 is the DNS server, the workstation has a dynamic address assigned by DHCPv6.

Resetting the IP stack on the workstation ("netsh int ip reset" and "netsh int ipv6 reset") doesn't help either.

Edit: It seems that the route isn't added on the client. Pinging link-local addresses works, from server to client and vice-versa. It also seems that I have to some router advertising with netsh. I'll look into this and will post a follow-up.

Edit 2: It seems that router advertisements are needed to publish a route on the network. Besides that, the "Managed" and "Otherstateful" flags need to be activated to enable DHCPv6.

I've now got IPv6 running, the only issue I face is that somehow the client gets 4 IPv6 addresses. 2 normal addresses, one temporary and one link-local. The 2 normal addresses probably means that one of the 2 is done via auto-configuration.

Is this the usual idea when using DHCPv6, or am I missing something?

Thanks in advance.

I have a server which basically acts as a router, but also offers some other services. The server has 3 interfaces, one leading to the main LAN and one to a server segment. Both these interfaces have an IPSec policy applied, but some servers on the server segment do not require IPSec or simply cannot function with IPSec enabled.

Once RRAS is configured, is the IPSec policy applied on the traffic that is routed through the server, or only on the traffic that is directed at the server, or does this depend on the policy that is used?

So if I want to connect to RRAS server, IPSec should work as expected, but if I want to connect to a server on the server segment from the LAN, the IPSec policy should not interfere and the traffic should be routed as usual. Is this possible?