Adam Davis's questions

In an existing Windows Server 2008 domain the users added by professional IT administrators automatically mapped drives when they logged onto any office machine.

The new users I've added, however, don't appear to be mapping drives as expected.

My process was as follows:

1. Start Active Directory Users and Computers
2. Under [mydomain] under User Accounts click Create a new user in the current container
3. Enter the user information and finish that dialog/wizard.
4. Right click the user, select Add to a group... and add them to the relevant folder groups and email distribution lists.

While the users have access to the paths, the drives are not mapped as is automatically done with existing accounts. I've gone through the Properties... and all the settings appear to be the same, including Connect client drives at logon being checked under the Environment tab.

I've followed a variety of instructions about modifying my sources.list trying to install Sun's JRE for Java 6 with no luck. The defining error I'm getting appears to be package sun-java6-jdk is virtual.

How do I install sun-java6-jre on Ubuntu 10.04?

We were dinged recently by our ISP when our server briefly saturated their line. Apparently they don't like bursty traffic. Who knew? Like many low cost hosts, they specify limits in terms of total monthly bandwidth, but they also 'reserve the right' to terminate network access for going over other unspecified limits, under the guise of protecting their customers.

I'm looking at modifying Apache mod_cband, but if that doesn't work, or isn't possible for whatever reason:

• What other programs/utilities can I install on a Windows 2003 virtual private server that allow me to control the bandwidth, both incoming and outgoing?

Google recently opened their Wave server for federation, meaning that anyone can download and deploy a Wave server.

How much 'server' is required to run Wave for a given number of active users?

I understand this is a difficult estimate to make, but I'm hoping that someone will have some metrics. If not now, then soon.

Java has increased in speed and performance, and I'm certain Google has been coding for speed, so I'm hoping that one can run a small wave server on a VPS, and a reasonably sized Wave community would only consume a single dedicated server.

But currently email servers can handle tens of thousands of users per mid-range server, and I'm worried that wave will require more than a single order of magnitude greater resources.

Apple has included HTTP Adaptive Bitrate Streaming in the iPhone OS 3.0, in particular Safari handles this automatically.

I'd like to play with this in a low cost manner, but I expect it'll require a custom HTTP server in the worst case, and interesting PHP/etc scripting in the best case.

But first I need to know what the protocol differences or standard is. HTTP is reasonably simple as a protocol, but adaptive bitrate means the file size is different, the chunk locations are different at different bitrates, etc. For instance, does the client tell the server anything special about the stream as it's downloading, or is it all handled on the server side?

Eliminating buffering pauses for the end user is very attractive for both live and pre-recorded video streams, and doing both over HTTP is even better given many networks and gov'ts are limiting non port 80 traffic.

• What are the technical details for HTTP adaptive bitrate streaming, especially Apple's implementation?

-Adam

If I allow multiple logins (ie, user "test" can login more than once) and a user doesn't log off, but leaves the sessions open, eventually the server will not allow any new logins (once the max is reached) - even for the user "test".

• How do I re-connect to a "dangling" session?

It appears that under some circumstances the remote desktop will only attempt to start new sessions, rather than reconnecting to the old sessions.

-Adam

I have a VPS which is limited to 2 concurrent TS sessions. At one point I was locked out from signing in, even though all my previous sessions were closed.

The provider indicated that the other sessions were still running, and they believe that I didn't actually "log off", so even if I exited the remote desktop client the server would still believe the connection was live.

I've never had this issue with a regular computer, when I kill the connection in any manner it always disconnects gracefully and doesn't lock me out.

• Does Windows 2003 have an issue with logging off properly when remote desktopping into it?
• How would I clear this error, or in other words if I need to can I kick off any current connections given that I have admin access?
• How do I check how many sessions are being used (assuming I can log in on a session)
• Can I load VNC as a 'backup' tool, or what other tools can I use to more fully protect myself from this issue?

-Adam

Vista now appears to support automatic wireless key distribution through USB disks. To see this in action, attempt to connect to a secure access point (that you don't have a key for) and when it asks for a key, insert a USB drive. It reports that the drive doesn't have the required information.

How do you create this file? It would be very handy for distribution of keys - in fact, one needn't ever give the key out - an administrator comes along, plugs in the drive, and then leaves. The users never have access to the key as it goes straight into the windows wireless configuration (although I imagine it could be extracted from windows).

Is it a simple text format, or more complicated, and how does one create this inside windows?

Is it a well enough known format that it could be used for MAC and Linux configuration as well?

Is this the same format that is used by XP (but isn't as 'automatic' as Vista's new support)?

-Adam

If NTP runs at 10:59:55, and fixes the time to 11:00:13, does chron run the 11:00:00, and *:00:00 tasks, or are they skipped?

Alternately, if NTP runs at 11:00:00 and fixes the time to 10:59:48, do these tasks run twice?

If they are skipped, what methods are used to fix this?

What about other schedulers (windows, etc)?

-Adam

First time VPS user, and still a beginner at Windows Server 2003.

Is the built in (simple) windows firewall considered sufficient if all I'm going to do is expose a few ports and block everything else, or do I need to consider another solution for an internet facing firewall?

-Adam

Windows Server 2003 (VPS)

Trying to install MySQL, but the service isn't starting - suspect the port is being blocked by the firewall.

One of my Google searches lead me to http://www.tek-tips.com/viewthread.cfm?qid=1117967&page=1 which seems to indicate I need to change my RRAS settings in order to access the firewall settings.

I've set up the RRAS so I can VPN into another network, so I'm reluctant to change them without understanding what the impact will be.

• Is there a way to access the windows firewall settings without modifying RRAS settings?
• How do I check if the firewall is even running?
• Is the RRAS change (disable NAT?) going to cause problems with my VPN connection?

-Adam

Most equipment is rated for a wide range of humidity (5 to 95% non-condensing, for instance).

However, what is the ideal humidity? Higher humidity carries heat away from equipment a little better, but may also be more corrosive, for instance.

There are lots of programs that small businesses (libraries, internet cafes, etc) can install on local computers to enforce their rules and terms of service when using free resources. These aren't ideal as they require constant updates, and can readily be circumvented. A lot of this stems from three intertwined requirements - preventing over-usage, decreasing liability (illegal internet activity, comply with gov't regulation, etc), while enabling the greatest level of internet access possible.

Assuming the computer cannot be trusted, a proxy appears to be a logical choice. A transparent proxy is nice in that it requires no configuration on the end machine, and would also apply to users that bring their own machines.

• What are good, free proxy servers that can be configured as transparent proxies?
• Is a proxy server good for metering and rate limiting usage, or is different software better for this?
• Are there free/open blacklists available for preventing the usual "bad" access (other proxies, child porn, etc)?

-Adam

A lot of time and columns are spent discussing securing a server from outside attacks. This is perfectly valid because it's easier for an attacker to use the Internet to break your server than it is for them to gain physical access.

However, some IT professionals gloss over the importance of physical server security. Many, if not most, of the most egregious breaches of security are performed from inside the organization.

• How do you protect your servers from users with on-site access who have no need to access the server or server room itself?

Is it just next to the IT manager's desk in a cubicle, or locked behind several doors with electronic card and biometric access?

Once someone has physical access to the servers, what protections are in place that prevent, or at least log, access to sensitive data they have no reasonable need to see?

Of course this will vary from organization to organization, and business need to business need, but even print servers have access to sensitive data (contracts and employee information) being printed, so there's more to this than might appear at first glance.

How does one fully evaluate a RAID configuration?

Pulling drives is one thing, but are there tools and techniques for more?

I've considered putting a nail through a running drive (powder actuated nailgun) to see what would happen, or simulating various electrical anomalies (shorts/opens in cable, power overloads and surges, etc).

What should be tested, and how?

-Adam

Under what conditions does one start to consider subnetting a network?

I'm looking for a few general rules of thumb, or triggers based on measurable metrics that make subnetting something that should be considered.

We've all seen good and bad examples of cable management.

• What are objective, measurable requirements that can be used in a policy to maintain cabling order in the rack/server room/data center?

I'm not looking for "Don't do spaghetti wiring!" but practical, objectively measurable policy that can be readily explained, followed, and inspected for whether it passes or fails the requirements in the policy.

Please avoid, "Don't do x, y, or z" - instead re-form the requirement as a "Do A, B, and C" where following the second requirement will eliminate the problems explained in the first requirement.