Large (i.e. 1 MB+) FTP uploads from newer client Windows OSes such as Windows Vista or Windows  7 running the FileZilla client (3.3.3) in passive mode, connecting to a Windows 2008 server would slow down, then hang/stall after a few seconds, eventually timing out.  In testing I found the issue occurred in both IIS/MS FTP/Microsoft FTP or FileZilla server running on the 2008 R2 box.  Here’s an example… on a Windows 7 FileZilla client, the transfer would hang at the EXACT same place EVERY time (in my case it hung exactly at 458,752 bytes):

Using a Mac or Windows XP FileZilla client did not produce a problem and the transfer always worked perfectly.

The solution was simple.  Enter this command line (must be run as admin) to disable the TCP autotuning mechanism on the server:

netsh int tcp set global autotuninglevel=disabled

Problem solved!  Big transfers – full speed ahead!  NOTE: I’m pretty sure this can be used on the client versus the server to work around the issue as well… please let me know in the comments.

Environment:

Server OS: Windows 2008 R2 Server
Client OS: Windows 7
FTP Server: IIS / MS FTP or FileZilla
FTP Client: FileZilla for Windows

Let’s take a look at your Sonicwall firewall rule (Advanced tab), namely the “TCP Connection Inactivity Timeout”:

See that “15″ minute value that it’s currently set to?  Change that to a larger number to allow your Sonicwall to keep the ports open during large transfers.

1. Open Disk Utility and use the New Image from Folder menu item to create an image. Ensure it is uncompressed and use the CD/DVD master option.In my experience this creates HFS+ masters which are no good in Windows.
2. Open the Terminal
3. Assuming your new image is called ~/Desktop/Master.cdr (the file is on your desktop and called Master.cdr) type:

This will create an ISO/Joliet .iso file.

In my example we will be paying Q2 2010.  Quicken automatically set up a sales tax tracking account where it automatically drops the sales tax from invoices I create.  I lump all my taxable sales into a category called “Sales” which you can see below.  This sales # is my taxable sales which I will need to calculate and reconcile with the tracking account.

Run a Profit and Loss Statement for the Tax Period you are paying by selecting “Business Reports” and “Profit and Loss Statement”

Change the report to Quarterly for the quarter you are paying tax on:

Here’s the report it generates:

In the case above there is $19,470 in non-taxable labor and $335 in taxable sales.  In Los Angeles there is an 8.25% state tax and a 1.5% district tax for a total tax of 9.75%.  This is what I have Quicken’s tax rate set to.  Doing the math with the above ($335 x 9.75%) I get $32.66 due to the California State Board of Equalization (our sales tax bureau in CA).  The California SBOE doesn’t like pennies so we just use whole numbers.  In my case $32.00 was payable for this quarter.

Now, Take a look at your Tax Collection account and make sure the numbers are equal, or at least close:

In my example you see that I owe $32.72 to the SBOE as calculated by Quicken ($32.66 was calculated based on the P&L).  The few pennies we are off here was due to my NOT adjusting the sales tax tracking account balance after paying my last return, which we will correct in the next step. To pay you sales tax, go to your checking account, and process your payment, selecting your “Sales Tax” account as the category (which treats it like a transfer):

In the transaction view for my “Sales Tax” account I simply add a blank debit transaction of $0.72 to zero out the balance and remove the pennies from the tracking account that I was not required to pay ($0.66), as well as the $0.06 that was there from NOT zeroing out the account last time I paid: $0.66 + $0.06 = $0.72.

Congratulations, you’ve paid your SALES TAX and are ready for the next quarter!

Creating a folder is easy.

To create a folder, touch and hold an app until it wiggles, drag it onto another app, and you’re done. You can repeat the process with more apps. For example, you can put all the games your daughter plays in one folder, all your news apps in another, and all your productivity apps in yet another. You can also create folders and organize apps using iTunes on your Mac or PC, then sync it all back to your iPhone.

Note that this tutorial does not address booting from MBR or dual booting Windows on the SAME hard disk.  I run Windows in a dual boot configuration, but I do it from an additional/physically separate hard drive.

1.  Connect your new hard disk to your Hackintosh via USB (enclosure) or directly via SATA.

2.  Load “Disk Utility” from Utilities on your Hackintosh and Partition your new Hard Drive.  I recommend creating a single partition formatted with “Mac OS Extended (Journaled)”.  Make sure “GUID Partition Table” is selected in “Options”.

3.  Clone your disk: Download and Install SuperDuper.  Run SuperDuper, selecting your source and destination drives.  Then hit “Copy Now”.  In my example below “Time Machine” would be the new, destination hard disk and “1TB” being the source disk.  The copy process will take a while.

4.  YOU ARE NOT DONE YET!  Now you must make the new disk bootable.  Download and Run the Chameleon RC2 Packaged Installer and point it to your new disk during the install process.  In this example, we’d be making the “Time Machine” disk bootable.  Requires NO terminal commands: SWEET!

Do you need or want a NEWER Chameleon release? Do this: Run step 4.  Then: Download the desired Chameleon release and replace the “boot” file in the root of your new drive (i.e. RC4 which has no packaged GUI installer).

5.  Remove your old hard disk and replace with the new disk.  Boot up and you should be good to go!  

NOTE: If you are using a USB enclosure for cloning, DO NOT TRY to boot your new installation from the USB drive.  IT WILL MOST LIKELY NOT WORK and you will get a “boot 1: error” in Chameleon!  You are not crazy and your clone is NOT BROKEN.  Simply remove the hard disk from the enclosure placing it directly connected to your system via SATA.

I recommend leaving your original “source” disk untouched and in a safe location for a week (just in case your new hard disk isn’t reliable or breaks, you can swap in your old disk).

Here’s my renovation using a 25 in Pegasus Montesol Granite replacement vanity top (the faucet is a Price Pfister Pasadena and the floors are Home Depot Castle Travertine):

After (with Flash)

After: Front View (with Flash)

After: Full View (with Flash)

After: Side View (No Flash)

After: Macro View (with Flash)

iPhone Pic

iPhone Pic (Close up)

Before Remodel (with Flash)

Here’s the quick and dirty:

1. Head over to Gravatar.com by clicking here: www.gravatar.com.
2. Create an account.
3. Link the e-mail address that you use on your WordPress blog/server to post/comment; then add your picture/logo/etc. to Gravatar.  Hint: if you use multiple accounts on your own WordPress site or other sites that use Gravatar, you can simply add the e-mails for those accounts to display the same Gravatar image that you uploaded.
   
4. In my experience your new Gravatar won’t appear without you adding a comment to any blog post.  I added one, saw my picture displayed, then immediately deleted the comment.  At that time all my existing comments (some from years/months prior) displayed my new Gravatar image.

Boom-shaka-laka!  You’re done.

My first dilemma was the Blackberry Controller Service wouldn’t start.  I got this in my controller log file:

[20000] (05/16 01:22:56.500):{0x147C} LogReceiver: Failed to bind a socket

That told me that there was a conflict with whatever port the controller was trying to grab.  After a little Googling I found that adding port 4070-4071 to the ReservedPorts TCP/IP registry key would fix this.  The Microsoft DNS service fights with BES for control of ephemeral ports which is a known situation.  Rebooted the computer and found that my devices were updating and the controller service started successfully.  However, my Application Server service kept shutting down every time I would attempt to access it, and could not enter the web administration interface.  I would ultimately get a connection timed out message.  I observed this in my system log:

Error: The BlackBerry Administration Service – Application Server service terminated unexpectedly.
It has done this 1 time(s).  The following corrective action will be taken in 60000 milliseconds: Restart the service.

And this in my BBAS logs:

(05/17 19:34:43:369):{main} [org.jboss.system.ServiceController] [WARN] Problem starting service jboss.cache:service=BASObjectVersioningCache
org.jgroups.ChannelException: failed to start protocol stack

I eventually found the fix here at the Blackberry support forums, which is to add an additional range of reserved ports for the Application Service:

Click Start, click Run, type regedit, and then click OK.
Locate and then click the following subkey:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters
Right-click ReservedPorts, and then click Modify.
Paste the following at the bottom.
48855-48858
45588-45588
49955-49955
45599-45599

Restart the server.

Once you go to the web admin, you will think its not going to work but hang tight, give it a few seconds and you will be surprised!!

Here’s what my ReservedPorts looks like after my two fixes (ignore the first three lines which were already there in my environment):

I don’t know how this person picked these ports (maybe spoke with support?)… but it DID indeed fix the problem for me.

My environment: Windows 2003 Server / Domain Controller / Exchange Server 2003 Standard

I developed the attached Windows batch file for use with VMware (I use ESXi 4) and ghettoVCB.  When used in conjuction with ghettoVCB’s log file, it will:

1. Detect if:
   a. The backup log doesn’t exist in the specified path (the backup didn’t run).
   b. There is a “Final status: ERROR” message in the log (error condition).
   c. The backup process ran with no errors.
   d. The backup log contained neither a good or bad backup indication (blank file handling).
2. After checking the above, the script will e-mail you the log file/result using BLAT.EXE mailer (enclosed).  It will clearly tell you in the subject if the backup was successful or not.  If your logfile does not exist (job never ran) it  sends an e-mail alerting you to this as well.
3. After e-mailing, the logs are rotated and the script will keep up to 6 backups of past logs should you need to trace an older backup.  The filenames will be “Backup1.log” – “Backup6.log”.  “Backup1.log” being the newest backup log rotated, and “Backup6.log” being the oldest.

Installation and Setup Instructions:

• Ensure that your ghettoVCB command line (cronjob) has the “-L” (capitalized for readability) parameter defined so that your backup log file has the same name and IS NOT date stamped by ghettoVCB.  The file MUST have the same name on every backup or my script will not work!  In my environment I use the filename “gvcb-backup.log”.
• Download the gvLog2Em batch file and BLAT.EXE mailer package here.  Extract to a folder/directory of your choosing.
• Edit the environment variables in gvlog2em.bat to match your environment.

Note: Make sure NOT to use quotations (even for long file paths) in defining variables.
Also make sure there are NO SPACES between the variable and the string “var=string”.

set blatpath=\\computer\share\path      <– The path to BLAT.EXE mailer / local or network
set logpath=\\computer\share\path      <– The path to your logfiles / local or network
set logfile=gvcb-backup.log  <– The name of your logfile

set temppath=%temp% <– Writable path to store temp mailbody text file.  Usually you can leave it as-is (I set it to the user temporary path).  New in 0.94. If the variable is omitted the script automatically sets it to the user temporary path.

set emailfrom=vmware@yourdomain.com  <– Backup log “From” email address.
set emailto=you@yourdomain.com  <– Backup log “To” email address.
set emailsmtp=mail.yourdomain.com  <– Your SMTP server
set emailport=25  <– Your SMTP port

set emailauth=False  <– Make this “True” or “TRUE’ or “true”  if you require SMTP AUTH.
(i.e. TrUe will not be found as a “true” condition in my code).
set emailauthu=username  <– SMTP auth username
set emailauthp=password  <– SMTP auth password

set esxiserver=myesxi.yourdomain.com  <– The name of your ESX(i) server, used in SUBJECT line (cosmetic).

• Make sure the Windows machine running this script has read, write and execute access to the logfile path.
• Test the operation of the script/batch in your environment, checking for errors in the console output.
• Once confirmed working, schedule a Windows task to run after your ghettoVCB backup.  (I use an XP virtual machine running on my ESXi server).

Tip: To enable logging of the batch file so you can troubleshoot your automated process, simply redirect it to a file when scheduling the automated task (after interactive testing), like this:

gvlog2em.bat > \\computer\share\path\gvlog2em.log

or really get nuts and put a date stamp at the end:

gvlog2em.bat > \\computer\share\path\gvlog2em_%date:~12,2%-%date:~4,2%-%date:~7,2%.log

Troubleshooting:

• “ERROR: Parameters missing or incorrect.  Check the configuration.”
  This means that one of the e-mail parameters is either missing or invalid.  Check your vars to ensure you have no spaces or quotation marks in your variables!
  
• “ERROR: Could not find the BLAT.EXE mailer at <blank>..”
  “ERROR: Without BLAT.EXE I can’t continue!”
  You left the blatpath variable empty or included spaces or quotes in the variable.
• “ERROR: Could not find the BLAT.EXE mailer at \\YOUR\network\path..”
  “ERROR: Without BLAT.EXE I can’t continue!”
  Could not find BLAT.EXE in the \\YOUR\network\path location variable.  Check that BLAT.EXE exists in the path you specified.  No trailing backspashes!

How I use it:

• I run a LOCAL cronjob on the ESXi server calling William Lam’s ghettoVCB.sh
• My rc.local injection boot-up command line for crontab looks like this:
  /bin/echo “0 0 * * 0 /vmfs/volumes/datastore1/Tools/ghettoVCB/ghettoVCB.sh -a -l /vmfs/volumes/readynasduo/VMBackups/Logs/gVCB-backup.log” >>/var/spool/cron/crontabs/root
  Note: 1. This is a single line (duh) and 2.  I specify the logfile name … no date stamping.
• I scheduled a Windows task to run the script 6 hours after my ghettoVCB backup (mine takes approx 2 hours to complete).

My environment:

• VMware ESXi 4.1 (new*)
• ghettoVCB.sh (latest) running locally within ESXi via rc.local injection cronjob.
• ghettoVCB backs up to a Netgear ReadyNAS Duo via NFS (datastore mounted via NFS)
  My backup happens once a week, Sunday at 0:00.
• Windows XP Professional VM on same server runs the task at 6:00am on Sunday.

Please let me know if you experience any issues or have improvements in mind!  I will continue to update this script as I receive feedback.

Changelog:

0.95 (2) – Fixed minor issue with feature introduced in 0.95, where the log would not be included if error or success conditions were not found.

0.95 – Handling of blank log files that do not indicate good or bad backup – will send an error message email.  Usually blank logs are due to ghettoVCB version issues.

0.94 – Better handling of non-existant/offline NFS systems/logpath shares.  Changed failure e-mail body bug that would not send an error e-mail if the logpath was not found.  New variable introduced “tempvar”.  See revision in the parameters section of the installation tutorial above.

0.93 – Minor cosmetic changes.

0.92 – Changed find behavior to look for final status error messages only, so that NFS sleep errors are not reported for successful backups.

0.91 – Added SMTP high priority flag to BLAT.EXE command line for unsuccessful backups.  Succesful backups are sent with the normal priority.  5/11/2010

0.90 – Initial Release  5/11/2010

Thanks to:

• William Lam for creating ghettoVCB.  The “Rolls-Royce” of ghetto VMware backup solutions.
• BLAT, one of the handiest utilities for Windows in my toolbox.. and all the folks who brought it to fruition!