Archive for the 'sysadmin' Category

X failures with ATI driver & Xinerama after upgrade

Published by
bigpresh
at November 9, 2007 in linux and sysadmin. 1 Comment

Arse - I updated my Arch Linux install last night, and X will no longer start using my previous config. I can start X using the vesa driver, so X itself isn’t broken, but when I try to start up with my multiple monitor setup (3 screens on 3 seperate ATI Radeon 7000 PCI cards) I get a crash, with the following backtrace:


Backtrace:
0: X(xf86SigHandler+0x7e) [0x80d9d5e]
1: [0xb7ef4420]
2: X(xf86RandR12SetRotations+0×6b) [0x80f58ab]
3: X(xf86CrtcScreenInit+0×9e) [0x80f173e]
4: /usr/lib/xorg/modules/drivers//radeon_drv.so(RADEONScreenInit+0×17fa) [0xb7adfcaa]
5: X(AddScreen+0×1fc) [0x80733dc]
6: X(InitOutput+0×21e) [0x80a769e]
7: X(main+0×296) [0x8073b66]
8: /lib/libc.so.6(__libc_start_main+0xe0) [0xb7cb9390]
9: X(FontFileCompleteXLFD+0×201) [0x80730d1]

How annoying.

I think the issue is down to changes in the X server for Xinerama (used to spread your display across multiple screens) to use RandR, which doesn’t seem to play well.

I’ll post updates here as/when I find a fix.

Using rankmirrors script for faster Arch Linux package downloads

Published by
bigpresh
at October 17, 2007 in linux and sysadmin. 3 Comments

If you use Arch Linux I’d strongly suggest using the snippet of code below to use rankmirrors to sort your pacman repository mirrors into an optimal order, to use the fastest servers first.


cd /etc/pacman.d
for repo in *; do
echo "Processing $repo..."
mv $repo $repo.b4.rankmirrors
rankmirrors -v $repo.b4.rankmirrors > $repo
done

I was getting an average of about 300-400K/sec; after doing this, I was reaching 10M/sec!

Joining the pool.ntp.org project

Published by
bigpresh
at September 25, 2007 in linux and sysadmin. 0 Comments

If you have your own server, why not consider joining the pool.ntp.org project to help provide accurate time to many client machines?

What is NTP?

NTP, or Network Time Protocol, is a method for computers to obtain the current time accurately in order to keep their clocks correct. Even when set accurately, the clock on most computers will eventually drift until it is no longer accurate - often made worse when the computer is being heavily loaded.

NTP allows a client to ask a time server to tell it the current time. The “main” timeservers have accurate timekeeping equipment connected (GPS receivers are common) to give them the correct time. Other timeservers simply keep in sync with several main servers, and accept requests from client machines.

What is the pool.ntp.org project?

The pool.ntp.org project is a big virtual cluster of timeservers striving to provide reliable easy to use NTP service for millions of clients without putting a strain on the big popular timeservers.

Continue reading ‘Joining the pool.ntp.org project’

New NTP server

Published by
bigpresh
at September 20, 2007 in linux, news and sysadmin. 2 Comments

Been meaning to do this for a while, but I’ve finally set up a new public NTP timeserver.

It’s a stratum-3 server, syncing to several decent stratum-2 servers, and answers to the name of time.preshweb.co.uk.

It’s entered the UK pool.ntp.org DNS pool, and I’m seeing quite a few requests already - currently my stats show 4.2 requests per second.

NTP is a damn useful tool to keep system clocks in sync and I’ve been using it for ages on all my boxen; it’s nice to be able to help out by sharing the load a little.

Read more info about NTP or about the pool.ntp.org project. If you have a server somewhere and are willing to help out, the project needs more pool servers to share the load - read more about joining.

Installing PCLinuxOS - what a breeze

Published by
bigpresh
at August 28, 2007 in linux and sysadmin. 0 Comments

I’m setting up a PC for my future father-in-law right now. I’m avoiding installing Windows as I hate dealing with it. I’m happy to help him to get to grips with using the machine and doing whatever he wants with it, but not so happy with having to provide Windows support (after all, I don’t use Windows, so I can hardly call myself an expert at it).

So, I needed a Linux distro that’s clean and simple that he should be able to just get on with. My friend Tony recommended PCLinuxOS as a suitable distro, so I thought I’d give it a spin. It boots as a live CD incredibly easily, auto-detecting everything. Once at a KDE desktop, it’s a usable system already. If you want to actually install to the HDD, just double-click the install icon on the desktop (yes, it has KDE set to require double-click for icons; no doubt useful to save confusing people moving over from Windows, but left me wondering why the hell it wasn’t doing anything). The process is simple, and the ability to sit here with a working system and a copy of Firefox to keep me amused whilst the installation progresses is actually quite cute :)

Continue reading ‘Installing PCLinuxOS - what a breeze’

Updated DNS propagation tracker

Published by
bigpresh
at August 15, 2007 in internet, perl, programming and sysadmin. 0 Comments

Updated my DNS propagation tracker, so that you can specify a set of nameservers to query rather than using a random set, or enter a domain to query the nameservers listed for that domain (useful if you want to check that all nameservers for your domain are giving the same answer).

Here it is: DNS propagation tracker

Damn hackers… :(

Published by
bigpresh
at August 7, 2007 in rant, security and sysadmin. 2 Comments

I got a phone call earlier today from Bytemark, who host a VPS server for me. I’ve had this box with them for a couple of years, and it’s been good.

However, it seems some idiot script-kiddie has used a vulnerability in some webapp to download and run a Perl script named udp.pl to do a UDP DoS against another server (something IRC related).

This had been happening since late Friday evening, and the excess bandwidth charge came to over £80 :( Now I don’t blame Bytemark for charging me really, they’ll be charged for the traffic so they have to pass it on - I blame the bastard who was abusing my server, and myself for (a) not having kept up to date enough, and (b) not noticing quicker (if I’d looked at my mrtg traffic graphs over the weekend I’d have instantly seem that something was amiss).

I’ve not been able to find any evidence as to how the attack came in, but I suspect it may have been WordPress - I hadn’t upgraded from 2.1.2, and apparently there was a remote code exploit in PHPMailer, which WP used, so WP was vulenerable to it. Other possibilities are Dokuwiki (which I’ve updated to latest version) and Gallery, which I’ve deleted.

On the plus side, I’m pretty sure they didn’t get actual access to the box beyond running their flooding script, although I’m still nervous trusting it, so I think I’ll have to move everything off it ASAP anyway.

In the meantime it’s behaving itself. Really not a good day today - slept through my alarm somehow and woke way, way late for work, then have this shite to deal with this evening… bah, hope tomorrow is better!

Lesson learned: make much more effort to stay up to date with all software; keep a much closer eye on the system; install Tripwire or similar; be more vigilant.

Finally implemented SPF

Published by
bigpresh
at July 23, 2007 in abuse and sysadmin. 0 Comments

Finally set up a proper SPF record this evening for my domain, hopefully that’ll help at least a little in the fight against spammers.

Continue reading ‘Finally implemented SPF’

Bash’s $_ variable (last argument)

Published by
bigpresh
at July 7, 2007 in linux, shell and sysadmin. 0 Comments

A few weeks ago, I found out about the incredibly useful Bash variable $_, which means “the last argument of the last command executed”.

It’s kind of similar to Perl’s $_ var in some ways, and can save a lot of typing.
Continue reading ‘Bash’s $_ variable (last argument)’

49MB mailbox size limit (Postfix & procmail)

Published by
bigpresh
at July 6, 2007 in linux and sysadmin. 10 Comments

Here for future reference more than anything else, and to help anyone who’s Googling for this.

I have my email delivered to a box at home running Postfix, and stored by procmail into mbox mailboxes.

I was finding that, whenever a mailbox file reached about 49MB, procmail would no longer deliver mail into that mailbox, instead dumping it into my main system mailbox.
Continue reading ‘49MB mailbox size limit (Postfix & procmail)’

get some tasty email addresses here