For reference, my cron line is */1 * * * * set_volume_mode Download which runs the command every minute. This may be a little un-necessary, but the script runs in a second or less and i’ve not seen any negative impact on the NAS.

As usual, if you follow these instructions, I take no responsibility for damage caused to persons, property, etc. You follow my instructions completely at your own risk. Bla bla bla :-p.

1. Login to AWS -> EC2
2. Shut down your EC2 instance
3. Click Snapshots on the left hand side and create a snapshot of your main root ebs volume
4. Once completed, click volumes on the left and create a new volume (with the new disk size and the snapshot you just created)
5. On the volumes page, make a note of the disk location (probably /dev/sda1)
6. Detach the smaller volume
7. Attach the larger volume (using the the ref noted above (probably /dev/sda1))
8. Boot the EC2 instance
9. Now login to the EC2 instance. If you run df -h, you will see the disk still thinks it is the same size as before. Make a note of the Filesystem (probably /dev/xvda1)
10. Run resize2fs -p /dev/xvda1 where /dev/xvda1 is the Filesystem noted above
11. run df -h. You should now see that the main partition is the size of the new volume you created.

As always, I accept no responsibility for damage caused, loss of data, loss of anything. You use these instructions at your own risk.

Elastix Logo

I’ve gone rather mad with VoIP configurations since I decided to move to using my own PBX for my personal and business telephony. My choice of installation was Elastix, mostly because it looks nice, but it works really well. These details will work with pretty much any other Asterisk PBX such as Trixbox though. Right, so let’s setup the SIP trunk. The details you need to set are as follows:

Maximum Channels: 2

PEER Details:
allow=alaw&ulaw&G.726&gsm&g729&g723&g777
canreinvite=no
context=from-trunk
disallow=all
dtmfmode=rfc2833
fromdomain=sip.voipcheap.com
fromuser=USERNAME
host=sip.voipcheap.com
insecure=very
nat=YES
port=5060
qualify=yes
secret=PASSWORD
srvlookup=yes
type=peer
username=USERNAME

User Context: leave blank

User Details: clear so it’s blank

Register String: USERNAME:PASSWORD@sip.voipcheap.com/USERNAME

So long as you have replaced all of the USERNAME’s and PASSWORDS with your VoipCheap ones, you should be laughing. Remember that to dial out, you will need to setup and outbound route!

UPDATE wp_posts SET post_content = REPLACE (post_content, 'OLD_DOMAIN', 'NEW_DOMAIN');

If you are running on a WP Multi-site installation, you’ll need to replace “wp_posts” with something like “wp_SITEID_posts” (e.g. wp_19_posts).

I hope someone will find this useful. Remember to always backup your database before you run any updates or changes just in case it goes wrong

sh-3.2# service sshd start
Starting sshd: Could not load host key: /etc/ssh/ssh_host_rsa_key
Could not load host key: /etc/ssh/ssh_host_dsa_key
Disabling protocol version 2. Could not load host key
sshd: no hostkeys available — exiting.
[FAILED]

The fix is simple, just run the following commands, then try and start it again and it should work

ssh-keygen -t rsa1 -f /etc/ssh/ssh_host_key -N “”
ssh-keygen -t dsa -f /etc/ssh/ssh_host_dsa_key -N “”
ssh-keygen -t rsa -f /etc/ssh/ssh_host_rsa_key -N “”

MySQL

Recently, A project that I was working on required that I import a fairly large BCP file containing around 260k entries. I spent quite a while trying to work out the best way to import it directly into my MySQL database with little or no pre-processing required. The answer turned out to be quite simple, I created a private directory (as the file contained some confidential data) on the server that was hosting MySQL, I then uploaded the 100mb bcp file to it and also created an SQL file with the following content:

LOAD DATA LOCAL INFILE 'BCP_FILE' INTO TABLE TABLE_NAME
FIELDS TERMINATED BY '@**@'
LINES TERMINATED BY '*@@*'
(TABLE_FIELD_1, TABLE_FIELD_2, TABLE_FIELD_N)

You then replace BCP_FILE, TABLE_Name and TABLE_FIELD_* with your details (BCP_FILE with the BCP file’s name, TABLE_NAME with the name of the table you want to import the data into and finally replace TABLE_FIELD_* with the fields in your table which match those in your BCP file). Once you have done this, save the file as something like bcp_import.sql. Just as a quick sanity check, you should now have a directory with two files in it, one bcp file and one sql file. Now for the fun part . Login to the server via SSH (or terminal if you are local), then login to mysql with:

mysql -u YOUR_USERNAME -p

You will then be prompted to enter your password. Next, select the database with the table that you want to import into:

use DATABASE_NAME;

Finally, run your import file with:

\.bcp_import.sql

That’s it, with a bit of luck, the data should now be in your database table. Be patient with it though if you are importing a large BCP (or have a very slow machine).

As usual, comments, suggestions, problems, thanks below

Google Chrome

Over the past few months, I’ve noticed that Google Chrome’s performance has been rapidly going down hill with crashes, freezing and generally being slow. It’s also been failing to update itself completely properly. Today I decided to just uninstall it then re-install it. After uninstalling it, I went to go and delete all of the old Application Data, where, much to my suprise, the Chrome directory was 17.8GB with 126k files. Digging a little deeper, the vast majority of this data was in the “JumpListIcons” and “JumpListIconsOld” directories which are found in C:\Users\<username>\AppData\Local\Google\Chrome\User Data\Default. Screenshot below of the 17.8GB folder, I have to say it’s pretty impressive! You can click on it to see it full size

Large Chrome Application Data Directory

I have now run with my newly installed Chrome for about 5 hours, and it looks like the JumpListIcons directory is under control. Fingers crossed that is problem solved!

As usual, if you have any questions, comments or anything else, the commnets box is below

/tmp/minisuhosin exploit

Today I discovered a random process running on one of my servers with the process name minisuhosin (/bin/sh minisuhosin). I believe this to be a new (past few days) edge exploit as there is very little information on this on google. Infact, the only other info I could find on it was this post here.

The first thing that I would like to state is that in the case that I dealt with, the account had both osCommerce and WordPress installed, although the exploit only seemed to be present in the WordPress directory. The contents of the .htaccess file was as follows

AddType application/x-httpd-php .php .phtml .php3 .php4 .php5 .htm .html
php_value auto_prepend_file /tmp/25454b22bf39c75795851f39d5e347c4

As the above mentioned blog post confirms, the file claims to be mini suhosin, and indeed this is not the case. From what I could tell, there was a single parent process which was constantly spawning child processes of the same name and then killing them. Whilst I am not aware of the full extent of the exploit (I will need some more time), the intent of the script only seems to target the website which was exploited.

It appears to try and load some javascript which returns some webpage containing some junk, most of which my firewall blocked (a java applet and a pdf). For reference, I have put this script here. It should allow you to tell if you have been subjected to the same exploit. On the basis that both myself and the site linked above, the filename was the same, I would look out for the “/tmp/25454b22bf39c75795851f39d5e347c4″ file.

In terms of prevention. Make sure that all of your scripts are up-to-date and that your /tmp directory is mounted with noexec.

If you have any insight into this exploit, please share your knowledge in the comments box below.

SBSettings iOS4

I’m a big fan of the SBSettings application for jailbroken iDevices (Available via the BigBoss repo on Cydia). After upgrading to firmware version 4.0.1 (using jailbreakme.com) and installing SBSettings via cydia, each time the installation completed and I clicked the respring button, the springboard would just crash and I would end up in safe mode. In my case, the problem was that some of the old toggles for SBSettings had been brought over in my iPhone backup. The solution was to SSH into my iPhone 3Gs, navigate to /var/mobile/Library/SBSettings/Toggles, then type ls -l to list all of the available toggles (with the dates they were added). The majority of them will be listed as a single date (the date which you installed SBSettings last), in my case it was “Aug 21″. Remove all folders which do not have this date (they were Toggles which were probably only designed for firmware 3.x, so were crashing 4.x in my case). I removed UAFaker, Rotation, EDGE, Push and Tether; I never used the EDGE, Push or Tether toggles anyway, so it was no big deal for me.

I hope that this helps someone. If you have any comments, suggestions or just want to say thanks (:-)), please leave a message below

As usual, I accept no responsibility for any damages caused, do it at your own risk. I’ve performed this on my own iPhone 3Gs on iOS4.0.1 and it worked, this is no guarantee for any other hardware/firmware combination (although it should work)!