VPS.NET Cloud Community: SteveSRS - Viewing Profile - VPS.NET Cloud Community

Hi,

Recently found out my server was vurnable:
http://httpd.apache....E-2011-3192.txt

I now fixed it with some extra rules in my htaccess file;
RequestHeader unset Range
RequestHeader unset Request-Range

And it works however I don't feel this is best solution and its not server wide, I'm actually not sure if adding these lines to apache2.conf would help.
An other fixed I tried (using mod_rewrite) didn't work when I put them in apache2.conf.

I run debian squeeze, and it says patch was already available but I ran:
apt-get update + upgrade but that did not solve any problems for me

Debian version 6.03
apache version: Apache/2.2.16 (Debian)

Anybody knows steps for a better solution?

Thanks

P.s. if you haven't checked your server I would advise you to do so.. tool is here:
http://seclists.org/...re/2011/Aug/175

Hi,

With help of on-demand support (great job guys) I know have a 3 server Mysql-Galera cluster (awesome!) however due all the work done we ran into some bugs (inc debian garbage collector bug) and at one time due to some errors my log files filled up as crazy (limited size of logs files now). However because of all this I'm more regularly checking disk usage on my servers. Now today I did a df -h and I got this:

/dev/xvda1 19G 4.2G 14G 24% /
tmpfs 374M 4.0K 374M 1% /lib/init/rw
udev 353M 80K 353M 1% /dev
tmpfs 374M 0 374M 0% /dev/shm
overflow 1.0M 1000K 24K 98% /tmp

That last line 'overflow' is completely new (yesterday that wasn't there) and well it doesn't look good. I'm not 100% sure but it looks to me like /tmp mount has reached its end. I actually already saw another post with someone with same problem however solution posted there was with use of cPanel which I don't use and I just want to be sure that a full /tmp is the problem.

I did solve a Mysql-Galera problem error 126 (faulty table) changing the my.cnf adding tmpfile = /somedir/ and setting it to other folder. So that made me also think the problem was the /tmp folder which is full.

Thanks for the advise.

Hi,

I recently had some work done by a freelancer (and way before by other freelancer) on my servers. However I was wondering, what do you guys check after job is done if they haven't left some stuff behind (knowingly or unknowingly). Also where should I check?

Couple questions:
1. I checked /root/.ssh/ for the keyfiles and found some ssh keys I did not recognized so deleted them, are they important to delete?

2. I did a list of users on my machines (debian) nothing weird there however I saw some usernames as 'irc', I don't use IRC for nothing can I just delete that?

3. I checked list of Mysql users (select * from mysql.user) and found I have 3 times 'root' and 3 times 'repl' user (repl user is user for master - master replication but don't get the 3 times). Why could that be? And is that something to worry about?

Where else should I be checking?

Thanks

I have 2 servers in the UK and US. Currently they are rsync-ed and have a master - slave setup where the UK server is master.
Now when the master goes offline (which happens too often) the whole eCommerce system isn't working anymore.

So i'm looking for a setup to solve this problem. So far I've found that a master-master setup would solve the problem partly as
the ecommerce functionality wouldn't be offline if UK server is offline neither if US server is offline, however as far as I've researched looks like master - master setup has some drawbacks and when one server goes down and then comes back online it will most likely have replication issues which need manual fixing.

So what I need is a master - master with fail-over solution, however this seems to be very difficult with just 2 servers.

Does anybody have some suggestion for a master - master + fail-over solution?

I found this which seems interesting:
https://code.google....nal_replication

I think this does what I need, however I could use a bit more of step-by-step on what to do (a howto would be nice).

Thanks.

Best regards,

P.s. also willing to pay for setting this up

Hi,

I ran into this;
http://www.securityg...rce-backdoored/

might be useful info from some of you out there.. However it also comes with a questions. I know I have OpenSSH installed. What should I be doing right now?

One of my servers (London C cloud) after the problems VPS had with the SAN is still having HUGE issues. I don't know if it is just coincidence or something more.. Support is helping me to look into the problem.

Also noticed some weird emails from different sources (email lists) with all kind of debian debug info and server details in the email..