JohannesH's questions

So I've just installed Icinga and Icinga-Web. And everything looked ok when first inspecting the UI's on http://[server]/icinga/ and http://[server]/icinga-web/...

I went on and added some new hosts, services, commands, etc. and reloaded the icinga service, without problems. Then I flushed the icinga-web cache and restarted the apache2 service, again no problems.

After reloading the configuration, all of the new hosts and services shows up just fine, in the old icinga UI, and the checks are running perfectly, well almost ;) (see below).

However, in the new icinga-web interface it's a completely different story. The icinga-web UI still shows the hosts from the initial configuration (a single 'localhost' host with 8 services). The UI also shows the 'default' instance as being 'down' in the header status, and that 'the data is X minutes old' (see below).

I've tried retracing my steps, by re-reading the icinga-web installation guide, but I can't figure out what I've done wrong or how to correct it. I've also tried to fiddle with the access.xml and databases.xml configuration files.

What am I doing wrong and how do I fix it?

In my experience both IIS Web Farm Framework and MSDeploy is doing a really poor job of managing configurations between servers.

While the idea behind the Web Farm Framework and MSDeploy is awesome, I don't think any of those two have done a good job of doing the task. Of all the setups where I've tried to use either Web Farm Framework or MSDeploy, it has failed, or at best, run very unstable. Usually I have to revert to manually configuring servers and mirroring files, which is less than optimal in larger setups.

So I wonder if there is any IIS configuration management on the market which actually does a decent job of synchronizing files, site configuration, certificates etc.

Is it ok to simply clone a vmware image of an Ubuntu server if I want an exact copy to run as a load balancing node? I realize I, at the very least, have to change the new machine's ip address and name, but are there any other things I should be aware of?

Clarification: The server is a standard installation running nginx and haproxy as a http loadbalancer and ssl proxy for some websites.

First of, I'm fairly new to this whole linux thing... So be gentle.

The setup:
I have a couple of Ubuntu servers... A pair of DNS servers running Bind, and a pair of servers running Nginx and HAProxy to load-balance requests to some back-end servers.

Now, the Bind servers pretty much keep each other up to date by doing zone transfers, although there are some settings in bind that I would like to share between them. However, The load-balancer servers are completely independent which is really bad. This is bad because I then have to change the configurations on both servers independently when I want to make config changes. This of course opens the servers up for mistakes where I accidentally configure the two servers differently, which will inevitably turn into "hard to debug" errors and bugs at some point.

The question
So my question is... How do I sync config changes between the servers, as well as automatically restart the services when their configuration changes?

we are currently thinking about running a medium size mail service.

Currently, dovecot seems to be the a good fit and we are about to setup a small test environment to check out some integration points with other systems. However, I was wondering if there are any good alternative mail servers (or services) that we haven't looked at yet (we've also looked at qmail). The alternatives doesn't exactly have to be free or open source but we do have some requirements.

Must haves:

• Must support delegation of user authentication to a centralized user registry.
• Must support between 50K and 150K users.
• Must support configuration of mailbox/outgoing/incoming mail size.
• Must support integration with an ingoing/outgoing spam/malware filter.

Optional (but much appreciated):

• Should have a API (preferably a cross platform one like a REST service or an open database) to allow querying of user stats like quota usage, spam status, unread mails etc., for use in other applications.
• Should have support for SSL-connectivity.

It seems as though Dovecot is a nice fit and I'm mostly just looking for confirmation that this is the right choice, or if not, what we should look at instead.

I have 2 web farm configurations, one with 2 member servers and one with 3 member servers. I have health monitoring set up on both farms and the monitoring tool reports all servers as being healthy. However after a while all the servers are marked as being "Unavailable" and "Healthy" in the "Monitoring and Management" screen (in the "Servers" screen they are all listed with "Yes" in the "Ready for Load Balancing" column).

Viewing the event log on both the web farm controller or any of farm servers doesn't reveal anything interesting. there are no warnings or errors in the period where the servers became unavailable. There are a couple of informational events about the worker process getting shut down due to inactivity but I don't hope this is the cause since that would mean that the farms will die during the night when the load is low.

Am I missing something?

EDIT: Btw, I think its very odd that the application pool shuts down on the servers since the health monitoring system is polling an aspx page on each server. Shouldn't that keep them going?

EDIT2: Now I've also experienced this problem with the RTW version of Web Farm Framework 2.

I have a problem that googling has turned up nothing, except this question on experts exchange which I borrowed most of the text from. :)

When I connect via Remote Desktop to a new Windows 2008 R2 server it takes 20-30s to get past the "Securing Remote Connection" message during the login.

If the password is wrong, it does this every time you attempt a login (ie its not a one time thing). However, after a successful login attempt the following logins to the same server goes faster.

Most servers runs on VMware here, but I don't know if that has anything to do with it.

We're in the process of installing a brand new Windows Server 2008 Web cluster and we would like to synchronize some files between the servers. The problem is that the DC in the domain is an old Windows Server 2003 Standard (NOT R2) which apparently doesn't contain some extension to the AD schema. Is it possible to upgrade the schema without upgrading the DC servers to R2?

When I try to create a Replication Group on the 2008 Server I get the following message:

---------------------------
Error
---------------------------
srv.XXXXXX.XX: The Active Directory Domain Services schema on domain controller
activedc07.srv.XXXXXX.XX cannot be read. This error might be caused by a schema
that has not been extended, or was extended improperly. See Help and Support
Center for information about extending the Active Directory Domain Services
schema. Schema version 30 is not supported.
---------------------------
OK   
---------------------------

A my current client we have a rather elaborate domain setup and I think it might be better to set up a trust relationship so the client domain is able to reference users and resources on the vendor domain. This will make administration of users much easier as well as making server administration much easier for the vendor.

So my question is... Is it? Easier, I mean. Is it easier to set up a trust relationship between two domains compared to managing, say 50, user accounts in 2 places?

Is it possible to reaname a TFS workspace? If it is, how is it done?

In my job I usually have to connect to several different servers with remote desktop. However, more often than not I will get the message "Terminal server has exceeded the maximum number of allowed connections", because my colleagues just closes the remote desktop window instead of logging out.

Usually I will call the admin and get him to release one of the idle sessions on the server but this method is getting real old, real fast. So I was wondering how I setup the server so that idle sessions will timeout after 5 minutes.