Welcome to MekTrix 4.0

There is a “media” share, which anybody can read, and the media user can write to. I’ve connected to this share, but can’t write to it. My first thought is that I must be connected anonymously, so I right click -> disconnect… and nothing happens. Right click -> disconnect again, “this drive is not connected”. Hmm, explorer bug? Well, if it’s not connected, then I shall attempt to connect. So I go to connect to network drive, click “connect with different login details”, enter details for the media user, click connect, and … “this drive is already connected”, and then a second error message immediately follows “you cannot connect to the same share with two different accounts” (What sort of stupid limitiation is that? :-|). Well, at least it confirms that I am logging in with two different accounts, and thus used to be anonymous and am now media. Or does it? I reboot to get rid of the phantom connection, and try to make a new connection with the media login. It works, but isn’t writable! Giving up on the client side, I check the server logs: “connect to service media initially as user nobody”. Gaaaah, goddamn it windows, I explicitly said “connect as the media user”!

Gaaaah.

</rant>

I don’t know if I’ve ranted on this last time, but it happened again with another site, so I shall rant again. For context, I’m the technical admin of a certain large site, which I won’t even name due to NSFWness — I’m just mentioning it here so that visitors from that site know who I am :) A brief timeline of events:

• 2 days ago, I move the database to a separate server, as MySQL grinds to a halt when the database is too big to fit in RAM.
• At around 11pm yesterday, the site suddenly dies, printing “500 Internal Server Error” in response to all requests
• Restarting the web server fixes it, but the old PHP processes don’t die, and the new ones lock up pretty quickly too
• PHP isn’t actually printing any error messages, it’s just locking up; LigHTTPD only prints “All PHP processes are busy, try again later”; MySQL logs show no error messages.
• It seems that the site is fine for as long as nobody tries to post new content.
• I try stopping and restarting the database, it says “Can’t stop database”
• I finally realise what happened — the disk with the database tables filled up, then MySQL crashed and died horribly, corrupting chunks of the tables. Then when PHP asks the database what’s up, and the database says “error”, it doesn’t print “there’s a database error”, it just dies horribly and locks everything up, causing ISE’s and using up all the ram until everything around it dies too.

In contrast, my preferred database server, Postgres, ran fine throughout the incident — read-only requests were served as normal, and if anything tried to add data it would be told “the disk is full”; no crashing, no corrupt data, no problem. Also, it seems to scale better, currently dealing with a database of 8 * the size of RAM quite happily.

It’s 7am, so I’m going to bed now; first thing I do when I wake up will be start work on giving this app Postgres support; then I can uninstall mysql since nothing else needs it \o/

It’s getting people other than myself contributing to it, such that I’ve had to learn how to set up subversion for multiple users :O

A problem I noticed, and google didn’t seem to have the answer for — when the client says “svn: CHECKOUT of ‘[filename]’: 403 Forbidden ([host])”, and the apache error log says “Access denied: ‘[user]’ CHECKOUT”, note that authz paths mustn’t end in a slash, unless it’s the root. eg, this works:

[shimmie2:/]
* = r
shish = rw

[shimmie2:/trunk/contrib/link_image]
Artanis = rw

But this doesn’t:

[shimmie2:/]
* = r
shish = rw

[shimmie2:/trunk/contrib/link_image/]
Artanis = rw

Blogging of my first days at uni has been delayed by problems connecting to the network… And the resulting conversation with tech support:

From: shish
To: tech support

Your SBS signup form for linux users [1] has typos[2] that make it
impossible to submit; the resulting page is a back trace of a
nullpointerexception[3]. I guessed the URL of the windows version of the
form and filled that one in instead, and it works fine (it even sent me
the correct download, CSA.sh). The fix is to go through the form and
give the elements their proper names (eg, change "username0" to
"username")

[1] http://reg-kent.org:8080/registration/register-linux.html

[2] It looks like someone copy & pasted the form in a WYSIWYG editor --
all the relevant elements had zeroes after them, which is a standard
way of keeping unique element IDs after copy & pasting...

[3] I would guess somebody did something like form["username"].value,
without first making sure that the "username" element was POSTed :-/

A rather simple problem, and I even took the time to work out why things were broken, and explain how to fix them in easy to understand terms. The result?

From: Tech support
To: Shish

Thanks for your e-mail.

You need to have Java turned on in order to see the page correctly and submit it.

Gah, a boilerplate response, and an incorrect one at that, on a variety of levels — there is no java on the page, so enabling it won’t make any difference. I would assume he meant javascript (something totally different, the name is coincidence), but that is inappropriate too, since the problem is nothing to do with the script (which works fine in my already javascript enabled browser).

So I reply again, in even simpler terms, and explain why the previous response was wrong:

No I don't; there is no java on that page. There is javascript (which
is a different thing altogether), and I do have that enabled. I know my
browser works fine since I can use the version of the page designed for
windows users.

To repeat: The problem is that the linux version of the page has errors
in its code that cause the values entered to be given the wrong names.
For example, the linux username box is called "username0", and doesn't
work, whereas the windows and mac pages have called the box "username"
and they work fine. If I download the registration page and change the
name of the box myself, that works too.

It's a very simple problem with a very simple solution; you could
probably fix it in less time than it would take to send another
useless boilerplate message :P

A response?

Thanks for your e-mail.

You need to have Java turned on in order to see the page correctly and submit it.

Argh.

For anyone with the same problems I had…

PEAR from debian stable is too old and won’t upgrade: download 1.3.5 manually, run “pear upgrade PEAR-1.3.5.tgz”; download the most recent version and run “pear upgrade PEAR-1.4.8.tgz” (in my case). You should then be able to let PEAR upgrade itself from then on.

MySQL extension is installed, but you still get the warning warning: pear/MDB2_Driver_mysql requires PHP extension "mysql": Make sure “module=mysql.so” is uncommented in /etc/php4/cli/php.ini.

After 8 hours of being under full load, chai (my linux router) panicked, taking the network offline for 3 hours. There’s nothing *obviosuly* wrong with the hardware, so I assume it a linux problem. I would have investigated properly, but the stack trace only showed the addresses of functions, not names…

Seeing as one error per 4 years is unacceptably high (from debian stable, of all distros!), I’ve since started looking into the BSDs; particularly m0n0wall.

After 4 months uptime, the server goes down for it’s first ever (?) voluntary reboot, with an extra stick of ram~

Ho hum, at least this shows the automatic recovery systems work – I was back online and serving before I’d noticed PuTTY had disconnected~

I’m also back on the front page of hits for “Shish”, above MSDN \o/

So I have a list of links; each has a bottom margin of 16px. When I hover over them in IE the bottom margins disappear. Setting the A:visited property makes things work, by disabling the :hover. Setting the width of the links also stops the jittering…

On the day I leave, the modem dies -__- Two week downtime, new record…