Eddie's questions

My desktop is part of the company domain and receives its IP address and DNS search list via DHCP. That's fine, of course. Is there a way that I can append additional entries to the end of the DNS search list that I got via DHCP?

When I bring up the NIC adapter settings for TCP/IP v4, of course it is set to obtain IP address and DNS server address automatically. Opening "Advanced" and clicking on the DNS tab, the DNS search list configuration is greyed out.

Is there a command or script I can run each time I boot to append to the end of the DNS search list that I get via DHCP?

My goal is to avoid having to type out long fully-qualified names for additional domains I frequently access (in our labs) that the company doesn't want to add to the DNS search list that everyone gets via DHCP.

Edit: In Linux, the configuration to do what I am looking for is easy. You configure either the DHCP client or the network interface itself to append or prepend or supersede the search list you get via DHCP. For an example of how to do in Linux what I am trying to do in Windows, see https://superuser.com/a/264032/2505 or http://jurjenbokma.com/ApprenticesNotes/domain_search_override.html

Does the lack of answers mean that this trivial configuration in Linux is not possible in Windows 7?

My Linux server uses LVM because I replace the machine a year ago and installed from scratch. LVM has been a life saver and I love its flexibility. My Linux laptop, however, was last fully installed three years ago, and thus uses straight partitions.

Is it possible -- without a reinstall -- to convert an existing system to LVM? If so, what tools can do this job?

I expect that any filesystems to be shared with Windows on a dual-boot system cannot be LVM. If I am wrong, please let me know, but I would be surprised.

When logging into Windows, it says on that page that CTRL-ALT-DEL somehow makes Windows more secure. I have never been able to figure a mechanism where having to press some specific key combination before logging in makes the system more secure. I have never encountered a VMS, UNIX or related system that makes you press any key to log in -- except older terminal-based UNIXes where you press ENTER to get a login prompt.

How does having to press CTRL-ALT-DEL before logging in make Windows more secure?

Sometimes I need to simulate a WAN environment between my client and server. I mean, specifically, from most important to least:

• Total bandwidth limit,
• Occasional long (many tens of seconds) delays
• jitter

Preferably, the WAN simulator could act as a proxy -- I point my client at the WAN simulator and the WAN simulator talks to the server. I have spent a lot of time Googling to try to find something that will do this, and have found very very little.

Of course, I know there are various solutions I can implement by controlling the network topology and inserting devices in the middle, but IT controls the network. I'm just a developer.

What free or low cost tools do y'all know about that can help one simulate a WAN?

I have a Linux Fedora server running Bind 9.5.0 and DHCP 3.0.6. Bind is configured to allow updates (I'm only showing part of the config, but the reverse part is there too):

zone "myhome.org" IN {
    type master;
    file "named.myhome.org";
    allow-update { key "myrndckey"; };
    notify yes;
};

and DHCP is configured to provide updates to Bind:

zone myhome.org. {
    primary    192.169.1.1;
    key        "myrndckey";
} 

When a Windows DHCP client connects to my network, I can see it by nslookup (dig, etc). However, when a Linux DHCP client connects, while it successfully gets a DHCP lease, its name and address do not appear in DNS.

Does anyone have any ideas what could be missing from my configuration that would cuase this? Could it be something missing on my Linux client configuration?

A friend of mine using (I believe) qmail has it configured to reject, as SPAM, any incoming Email with a Date: header that is too far from real time. I can't figure out how to do this with sendmail. How do you do this?

An application running a lightly loaded Apache HTTPD 2.0 has occasionally had problems where one (or more?) of the Apache processes took 100% CPU. We currently run HTTPD 2.2, I we may have seen this with 2.2 as well. I'm not certain. In some cases, the CPU usage was such that it blocked all but console access to the Windows server hosting HTTPD. I have never been able to track down what can cause Apache to do this.

The environment is Apache HTTPD directly serving static content, using mod_rewrite but not much else custom configuration. HTTPD is talking to Apache Tomcat (5.x) via mod_jk (1.2.25).

Has anyone else encountered this and solved it? The workaround we installed is to limit each Apache HTTPD subprocess to a maximum number of requests with the following configuration:

MaxRequestsPerChild 1000

where because the application uses HTTP/1.1, this is really more than 1000 requests per child process and more like 100,000 requests per child process.

There are many VNCs out there. realvnc, tightvnc, and so on. Which one would you recommend, or would you recommend avoiding the VNC protocol altogether? (And if so, in favor of what?)

Let's say you have a web app and you distribute an embedded Apache HTTPD with that web app. It's difficult to manage customers that want to run SSL. The best way appears to be to let the customer manage installing SSL certificates on the embedded Apache -- or to recommend the customer use a front-end SSL accelerator card or pizza box.

Are there better ways of doing this?