wehuberconsultingllc.com

This week I decided to update my Subversion components and it did not go well. The first surprise was after updating my version of TortoiseSVN  to the 1.7 version, my Subversion icons disappeared. The fix was simple but it reminded me that it had been a long time since I updated the Svnserve server I was using from CollabNet. This would also be a nice time to add HTTP access to the repository. So I downloaded the CollabNet Subversion Edge Server and tried to get it to work with my existing repository. I ran into several problems(e.g. viewvc exception).  While researching the solutions I decided an easier solution to my problems was to install the VisualSVN Server. So I uninstalled the Edge Server, installed VisualSVN Server, and imported my repository. This worked for HTTP access but I still wanted Svnserve access since I had existing folders checked out via this method. The VisualSVN bin directory has the current version of svnserve so I needed the instructions for installing it as a windows service.  The instructions I found, Svnserve Based Server, worked for me but not until I stopped fixing things I should not be fixing in the “unusual command line” format.

Note the rather unusual command line format used by sc. In the key= value pairs there must be no space between the key and the = but there must be a space before the value.

So now I have a relatively simple Subversion server solution that works for both svnserve and http access and it should be easy to maintain.

135 terabyte, 4U server for $7,384 is pretty cool!

It’s been over a year since Backblaze revealed the designs of our first generation (67 terabyte) storage pod. During that time, we’ve remained focused on our mission to provide an unlimited online backup service for $5 per month. To maintain profitability, we continue to avoid overpriced commercial solutions, and we now build the Backblaze Storage Pod 2.0: a 135-terabyte, 4U server for $7,384. It’s double the storage and twice the performance—at lower cost than the original.

Backblaze Blog » Petabytes on a Budget v2.0:Revealing More Secrets

Last Friday I had a problem seeing my shared network drive. Today I figured out that Netbios over TCPIP was disabled. I am not sure how this happen.

This morning I tried to start up my laptop at work. It had just gotten beyond the Lenovo logo when it said it had a Fan Error and shut down. My boss came in and took the battery out. He waited 20 seconds and then put it back in. It started up fine. Ouch!

Recently I have been interested in possibly converting my existing VPC and VMware images to VirtualBox. I have one image running under VPC and several images running under VMware server and ESXi. Although VMware has been an adequate platform the idea of running VirtualBox under Ubuntu server is appealing.

My first test was to try and convert my VPC image. It did not go well. I kept got a blue screen of death and wasted a lot of time. After soon research I realized that I did not setup the image correctly for migration.  Someday when I have a lot more time, I will try again .

My second test was to convert a mail server image that sends out our newsletter. This mail server has a history of being pretty compute intensive since we use DKIM to sign every email. From previous experiments I determined that I could shorten the run time by setting up the mail server as a multi-processor image. The host was on an old dual processor with 4 GB of RAM and under VMware Server  2.0 I could take advantage of both processors. However when I converted the image to run under VirtualBox I could take advantage of only one CPU. Alas the processors do not support VT-x so VirtualBox restricts you to one processor. I ran one newsletter through the VirtualBox version and it took about twice as long. Oh well! Back to the drawing board.

Recently I started getting some Flash display problems. The Flash graphics was appearing in the upper left corner of the screen and flashing intermittently. I believe these problems started occurring when I upgraded to the most recent version, 10.3.181.14. Here is what I tried and what worked for me.

1. I tried reinstalling the current version and the symptoms remained.
2. I tried a forced uninstall and a manual install of the current version. The symptoms remained.
3. I uninstalled the current version and installed the previous version, 10.2.159.1. I got the previous version from Archived Flash Player versions page. Now my graphics are working properly.

I had some funky display show up when I went to finance.yahoo.com so I ran a antispyware check using  http://www.superantispyware.com/index.html. Since this program takes a long time I ran the Microsoft Safety Scanner, too. Microsoft did not find anything and SUPERAntiSpyware found a false positive on a file included in QuickBooks SDK 10, tiny.exe.

Do you think your PC has a virus?

The Microsoft Safety Scanner is a free downloadable security tool that provides on-demand scanning and helps remove viruses, spyware, and other malicious software. It works with your existing antivirus software.

Microsoft Safety Scanner – Remove Spyware, Malware, Viruses Free

Last week I decided to update to the latest recent on my VMware version of the Ubuntu Desktop and it did not go well. During the upgrade I got a disk space error message with non-displayable characters. It was a curious message since I thought I had enough disk space and it did not stop the process. It was when I rebooted I encountered the major problem. I would like to tell you more about the error message but the screen clears and I am left with a GRUB prompt. That’s not nice!

So I did some research and booted using the instructions from this page, Express Boot to the Most Recent Kernel. Everything came up nicely except for my windows were missing the title bar when I came in via the NX client. To fix the booting problem, I reinstalled from the LiveCD using the simplest method. To fix the missing title bar I installed the latest version of NX client from NoMachine NX. I also got the missing title bar when I used the VMware console window to log in. In that case I just specified that I wanted to use the Classic interface when I logged in.

Over the last couple of weeks I have been working on an automated pricing program. This was not one of the algorithms I was looking at.

A few weeks ago a postdoc in my lab logged on to Amazon to buy the lab an extra copy of Peter Lawrence’s The Making of a Fly – a classic work in developmental biology that we – and most other Drosophila developmental biologists – consult regularly. The book, published in 1992, is out of print. But Amazon listed 17 copies for sale: 15 used from $35.54, and 2 new from $1,730,045.91 (+$3.99 shipping).

Amazon’s $23,698,655.93 book about flies

Windows Update was failing to install several updates Office 2003( KB2502786, KB2509503, KB2522981, and KB2464588). I eventually found the answer. I opted to manually update the registry. I only had to apply the first registry change described in KB2258121 for the updates to work.

This article (KB2258121) provides a fix for a problem in which you receive error code 0×80070646, error code 646, or error code 1606. This update issue occurs when you install Office updates by using Windows Update or Microsoft Update. For detailed information, see the “Symptoms,” “Cause,” and “Resolution” sections of this article.

Error code "0×80070646," "646," or "1606" running Windows Update

I have a working configuration for Advanced Web Statistics 6.9 (build 1.919).  When I upgraded to the latest release (Advanced Web Statistics 7.0 (build 1.971), it did not work. I ran into this problem in 2009 and backed off.  Now it is time to commit some brain power to fixing the problem.

Problem

When I tried to process a log file no statistical data was being collected. No error messages were generated even though many new records were processed. The data file remained very small.

Analysis

I set up a test environment in which I could process the data with the new version and debugging turned on. To turn debugging on I changed DebugMessages from 0 to 1 in the configuration file and added "-debug=10" parameter to awstats.pl command used to process the log file. This gave me all of the debug messages printed to the screen. I started processing a log file and interrupted it after it had processed a few lines by hitting Ctrl-C. In the debug output I could see that 7.0 version was having problems with the log format. Some values were parsed correctly and others looked like they were offset by one variable. In my configuration I was using a custom logformat which used a combination of IIS column names and AWSTATS keywords. I don’t remember how I created it but I suspect I found an example on the Internet. It worked for the last two years but now it seemed to have a problem with cs-username(%logname).

Solution

My solution was to convert the IIS column names to the corresponding AWSTATS keywords. Even after using the AWSTATS keywords it was still not parsing correctly. To get the parsing to work correctly, I had to change %logname to %other. In my log files, the value for cs-username(%logname) is a single dash.   Obviously it is not important or necessary. When I look at the code, line 9027, I see a comment that says the regular expression code allows spaces to support Lotus Notes. As a result it puts two fields into one variable. Naturally the parsing for %other works correctly. After I knew what the problem and answer was, I could find the answer on the Internet. Here is someone else who experienced the same problem, http://www.internetofficer.com/forum/awstats-iis-installation-and-configuration/awstats-wont-recognize-logname-cs-username-field/. It sure would have been nice if Awstats re-mapped cs-username to %other rather than %logname.

From the downloaded ISO it took me 1 hour and 37 minutes to install SP1 and reboot. In fact it took me 21 minutes to shutdown and reboot since it had a lot of updating it was doing during the shutdown and startup.

I keep our web site source files, stored procedures, and assorted programs in a source control management(SCM) program called Subversion. Since I am a small shop I track my trouble tickets using Spiceworks. My normal work flow is to enter the trouble ticket into Spiceworks and then start working on the problem. When I am finished with the problem I commit the files to the SCM and close the trouble ticket. To make life a little easier with future problems, I include a reference to the Spiceworks trouble ticket in the Subversion commit message. This could be a text version of trouble ticket number but I prefer a little nicer web interface so I implemented the bugtraq interface. This gives me a click-able link in the log that will bring up the original trouble ticket in a web browser window. This can be handy when you are troubleshooting a file that has some modification history. Implementing the bugtraq interface is pretty simple. You add two properties on your subversion repository, bugtraq:message and bugtraq:url. I set my parameters on the trunk folder properties. I set the message parameter to "Issue : %BUGID%" and the url parameter to "http://svrdev1:9675/tickets/list/all_tickets#ticket_id_%BUGID%". If you are planning to copy my parameters you should omit the double quotes and change the "svrdev1:9675" to your Spiceworks host name and port number.

If your SCM is Mercurial you can implement the bugtraq interface using these instructions, http://www.offroadcode.com/2010/10/5/better-issue-tracking-with-tortoisehg-and-kiln.aspx. The instructions are for a different SCM, Kiln, but these instructions work for any SCM you access via a url. In this case the issue link in the issue tracker changed to "http://svrdev1:9675/tickets/list/all_tickets#ticket_id_{1}".  There are several valid formats you could use. Just look at the regular expression used in setting it up for details. I used "(?i)(?:case|bugzid|Issue|#)\s*:?\s*(\d+)" for my Issue Regex. When you browse the log you should find the "Issue : 619" is an underlined, click-able link to your SCM trouble ticket.

I bought two of these this week. My Blackberry was working fine but my wife’s phone was having problems. The price was right but I cannot find the $50 mail in rebate form. I had a surprisingly difficult time ordering the phones via the t-mobile web site using Internet Explorer. I ended up using Google Chrome to get the pop up windows to show up.

I also had a difficult time changing my handset on their my.t-mobile site. It kept changing back to the Blackberry even though that phone did not have the SIM chip anymore! When the site changed my phone device to the Blackberry, it would complain that I had the wrong data plan for the phone. I had to reset the phone almost every time I logged into my.t-mobile.com.

If you’re looking to snag a T-Mobile G2 (check out our review) on the cheap, it’s now available for the low, low price of free through Jan. 20. With it you get great keyboard, fast processor and that crazy locked-down hardware that took way too much time and effort to crack. (But cracked it was. Huzzah!) Anyhoo, you’ve got a couple days left to pull the trigger. Get it at the source link. [T-Mobile]

T-Mobile G2 free through Jan. 20 posted originally by Android Central

Sponsored by Android Cases and Accessories

T-Mobile G2 free through Jan. 20
Phil Nickinson
Tue, 18 Jan 2011 01:45:01 GMT

I think I was bored on a slow Friday afternoon so I started updating non-critical software.   Somehow I found a portable version of Perl, Strawberry Perl, and an IDE for Perl called Padre. An IDE is a help for someone like me who does not use Perl very often. Since I have a love-hate affair with the Windows registry, I also prefer portable applications. Since all of my portable applications are Portableapps.com applications, I prefer portable applications that fit in this framework. So I found a “portable” Strawberry Perl. However it was not in the Portable Apps framework. I found an installable version of Padre but not a portable version. I am guessing that there is a Perl/Padre version issue since they only build Padre for Perl 5.10. The “portable” version of Perl I was looking at is 5.12. So I decided to ignore the problems and create my first Portable Application. If it doesn’t work I won’t use it. Here are my steps for creating a Portableapps version of Strawberry Perl with Padre(45.7 MB).

1. Download Portable Strawberry 5.12.
2. Create PortableApps folder structure for new application called StrawberryPerlPortable
3. Create StrawberryPerl Portable Application
   1. Install Portable Strawberry in "app" folder.
   2. Create appinfo.ini
   3. Create Launcher folder and StrawberryPerlPortable.ini pointing to portableshell.bat
   4. Create Portable application using Portableapps.com Launcher
4. Test
5. Install Padre using cpanp. Ignore errors.
6. Modify portableshell.bat to start padre rather than cmd.
7. Test example scripts
8. Final Test: Install my cidrmerge application.
   1. Install NetAddr::IP::Lite
   2. Install Net::CIDR::Lite
   3. Test cidrmerge.

I got a parse error when I upgraded from 0.14 to 0.15. I deleted the word “static” from the functions in the lightbox.php file as a temporary fix.

We use a web application to create purchase orders. We send purchase orders to our vendors by printing the web page to our fax server. Two days ago we sent out two faxes with large black blotches on the page. There were purchase orders that were sent out before and after these faxes that were fine. Something was different about these purchase orders. We found that we could duplicate the problem on different workstations. When I previewed the fax on my laptop I could see that it was already corrupt. If I printed the web page to the printer or created a PDF it was fine. My boss thought it might have to do with our recent roll-out of the new Windows 7 laptops. So I checked the HTML for odd characters and did not find anything. It was about this time I noticed that the two problem purchase orders were about the same size. One purchase order had 36 line items while the other purchase order had 37 line items. I also found out that I used Firefox or Google Chrome to print the web page, it created the fax correctly. This printing problem was specific to Internet Explorer. The final clue was found when I went to my old laptop and it displayed the faxes correctly. When I checked its page setup I noticed that it was using an A4 paper size rather than the standard Letter size. After a little fiddling with the page settings, I determined that when I changed the bottom margin on my new Windows 7 laptop to 0.5 or 1.0 inch the fax would be created correctly.

My boss remained unconvinced. Why did we not see this problem before? That was a good question. After a little investigating of our fax log I found a corrupted fax over a month ago.  Since this problem is rare and the subsequent faxes went out successfully, they resent the fax manually and did not bother to tell anyone of the problem.

In all of my years of being a system administrator I have never seen a machine as infected as I saw today at my sister-in-law’s house. When we turned on the machine, the fist sign of problems was that it did not go to the normal home page. When I tried to do a search the links on the results page did not work. Naturally my sister-in-law did not know what happened. Since the MacAfee software had expired and my sister-in-law was fussing about the cost of virus checking software, I downloaded a current version of Microsoft Essentials. I had to reboot the machine into safe mode with networking support before I could download the current virus and spyware definitions. Microsoft Essentials found a multitude of Trojans and worms. I cleaned the computer and rebooted several times. Finally I decided to perform a full scan before clicking on the “Clean” button. The full scan took a long time but the computer is now working as expected.

A big thanks goes out to the obligatorymoniker and his script for programmatically adding IP restrictions to IIS6. I was looking for a better script to add IP restrictions. My previous script added the restrictions one IP range at a time. This script was adequate for a small number of IP restrictions but recently I was asked to add IP restrictions for every country we do not ship to. We had credit card fraud transaction from one of these countries and the boss was mad. After using Perl and a CIDR to merge the adjacent networks, I still had over 18,000 IP ranges to deny. Using my old script I tried to add these IP ranges to our test system this took over an hour to load. Your script loads the ranges in a couple of seconds. Here is how I did this:

1. I went to http://www.countryipblocks.net/ to get the IP ranges I wanted to block. Beware these ranges include bogon networks(e.g. 192.168.0.0). The first time I applied the IP ranges I blocked myself out.
2. I used the perl script below to merge the networks.
3. I used the obligatorymoniker IP Security.vbs script to load the ranges. You will have to change the "IIS://localhost/smtpsvc/1" to the site you want to add the IP restrictions to.

use Net::CIDR::Lite;
use NetAddr::IP::Lite;
my $cidr = Net::CIDR::Lite->new;
# Disallow IPs

open (IPDISALLOW, "ip_disallow.txt") || die "couldn't open the file!";

while ($record = <IPDISALLOW>) {
 if (substr($record,0,1) != '#'){
#print $record;
 $cidr->add($record);
 }
}

close(IPDISALLOW);
#print "$_\n" for $cidr->list;
foreach ($cidr->list) {
my $ip = new NetAddr::IP::Lite $_;
#print "The address is ", $ip->addr, " with mask ", $ip->mask, "\n" ;
print $ip->addr, ",", $ip->mask,"$_\n";
} 

Recently I installed iRedmail at work so that we could include DKIM signatures in our newsletters. Every week we send out a newsletter to 96,000 former customers. It takes about 13 hours to send the newsletter. Yahoo is probably our most important email domain and they want us to implement DKIM. A couple of weeks ago we started seeing Yahoo limit our sending rate. Obviously they had a problem with something in our newsletter. So we re-analyzed the error codes we were getting during the newsletter mailing and implemented DKIM. The problem is fixed. Here is how I implemented this version of iRedMail.

I implemented a VMware version of iRedMail to sign newsletter emails using DKIM. I used Ubuntu 9 server version(optimized for VMware version) to build appliance.

1. The server works as a mail proxy in front of the SMTP server we use exclusively for the newsletter. It signs and relays the email to the existing SMTP server. I kept the existing SMTP server so that I could continue to use my existing procedures for parsing the log files to identify old/obsolete mailboxes.
2. I created iRedMail users in LDAP to relay local users to mailboxes on Exchange.
3. My primary bottleneck is still my mail transmission to the Internet speed, 2 per second. I can create newsletter emails at about 8 per second.
4. On an old Proliant DL350 G4 iRedMail consumes about 40% of the dual CPU computer for four hours.

Since I had experience installing iRedMail it went quickly. The biggest bug I had to fix was the AWStats/permissions problem on the mail.log file.