It has been my source for the SCO Linux fight since she is one of the
few rational sources of knowledgeable news.

But yesterday morning I couldn't connect with my home dial in. Waited
till I got to the office and my T1 line and still couldn't.

This morning tried again - same thing.

Did she get DOSed by the SCO scum? A cease and desist order? Server
crash?

Thanks
DD

On Thu, Mar 04, 2004 at 08:14:54AM -0600, DelphiDude wrote:
Hmm. Dunno. They're pinging, just not serving up any web pages.

--
GNU/Linux revenues last quarter: $1 Billion.
Apple revenues last quarter: $2 Billion.
micros~1 revenues last quarter: $4 Billion.
It's no longer a question of windows or GNU, it's a question of *Unix* or GNU.

Billy O'Connor wrote:

Nah, just slashdotted. It's slow, but hitting reload brings up the pages.

--
Change to leews to mail

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Lee Wei Shun wrote:

No, not slashdotted as the latest things haven't linked to them from
slashdot... just gone down under the sheer weight of "normal" traffic...
It's rather busy right now with all the latest griff going down.

- --
Twenty years from now you will be more disappointed by the things that you
didn't do than by the ones you did do. So throw off the bowlines. Sail away
from the safe harbor. Catch the trade winds in your sails. Explore. Dream.
Discover.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)

iD8DBQFAR1fnNclAUt2HMX8RAku2AJ96zMt1Cx/I0wMqNX5+oBtSiABucwCfXX8b
tBjHIeJDkr7ooSWxPbIFeoI=
=oeF+
-----END PGP SIGNATURE-----

DelphiDude said the following, on 03/04/04 09:14:
AFAIK, the problems are just due to very heavy traffic at the site. SCO
had announced they were going to announce [,,,] a law suit, and their
quarterly earnings announcement was also yesterday. And Slashdot had an
article about the first suit at 8:42 AM.

I got an E-mail from PJ at Groklaw yesterday morning, in which she said,
"Groklaw is so slashdotted, I can't get there to post the news."

Things seem better at the moment: it's still slow, but it does
eventually work.


--
Rich Gibbs
rgibbs@his.com

DelphiDude wrote:

She's using MySQL and MySQL scales like crap. So, when people reference her
site from slashdot, we'll wait forever and eventually get a MySQL error.

She should use PostgreSQL, it scales much better under heavy load.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

mlw wrote:

Riiiiight ... It couldn't in any way have been because of an
underpowered machine, it just *had* to be the database used. Because we
all know that even a Pentium 90 with 16 MB of RAM and an ISDN connection
will serve billions of hits a day, if only it uses PostgreSQL instead of
MySQL ... Sheesh ...

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFASJFhd1ZThqotgfgRAoZNAKDH3F98xp8BcL3Iglr4uw uikWrVdQCfSPk3
1KZCRIv0ddjIwniVfPoFUIs=
=ypOy
-----END PGP SIGNATURE-----
--
PeKaJe

Darth Vader:
The Force is strong with this one.

Peter Jensen wrote:

Well, if the machine came back with nothing, or didn't respond at all, I
would say that it was underpowered or having a problem. When it does
comeback with a MySQL error, I think it is fairly safe to conclude it is a
MySQL scaling issue. No?

Fearing a spontaneous XP reboot, mlw mumbled this incantation:

Gentlemen, restart your engines!

--
No, I won't fix your Windows computer!

Linønut wrote:

is really important, especially in the context of Linux advocacy.

I love the groklaw site. I think it is great, but because it uses MySQL, it
craps out under heavy load. I'm sure that the creators of the site said the
mantra of the MySQL supporter "it does what I need" or "It's good enough."
Well, my answer to these assertions is that, no it doesn't and no it isn't.

People should look long an hard are groklaw as a case study of MySQL under
load. Next time it is referenced by slashdot, what it crash and burn. It is
because MySQL does not scale. Period.

How the hell can we advocate how great Linux is, when the sites we reference
on a regular basis crash because of MySQL's poor design and inability to
scale?

mlw wrote:

No

So you have decided for yourself. Not for anyone else

Sure. No other explanation possible. Not even thinkable

Idiot

--
Try to be the best of whatever you are, even if what you are is
no good.

Peter Köhlmann wrote:

Then we have a difference of opinion.
It is my "professional" opinion.

high load it gets errors from the database. Give me an alternate
explanation.

You seem to like that word. You type it a lot. Is it some sort of self
descriptive sign-off or signature?

On 2004-03-05, mlw <mlw@nospam.no> wrote:
There are a number of possible problems, the configuration file and setup is
wrong, not enough mem on the server for the config file, db's not optimized,
not using indexes...etc.

Of course, it dosen't help to deliever 166k of single page dynamic content
from any database then format and display that data.

It also accounts for why it took 40 sec to display a page.

ken

ken wrote:

Well, there are a few facts that we can infer that do dispute your
assumptions:

(1) The site works normally under low load.

If we were talking about oracle, postgresql, or some real database with an
planning query execution, I could see that maybe something about the DB not
being optimized may make a difference. With MySQL, I'm doubtful.

Not using indexes? Well, I don't know the size of the tables in the
database, but it is pretty responsive under low load and I would tend to
think that it is indexed and the indexes are being used.

MySQL is a fairly simple database, one of the reasons people seem to like it
is because it is supposed to be simple to setup and operate.

That depends on the system they use to do this.

Not sure I agree. There are a number of reasons why it would be slow.

Hey, relax guys and gals. All I wanted to know if she was still up.

Besides, whatever you think about MySql vs Postgres, be glad she didn't
use MSAccess. Hmmm... I wonder how fast that would run under Wine???

But, while we are on the subject, it is my unproven opinion that MySql was
written in a secret Microsoft lab for the express purpose of killing
Linux. Kind of like the secret org that is turning Mblaster and cousins
loose on the world to kill Windoze:-)

Anyway, I am glad she is still around and access is much faster today.

DD

On 2004-03-05, mlw <mlw@nospam.no> wrote:
How do you know I didn't visit the site during a low load period.<g>

Regardless, I'm not a big fan of mysql. The site is using a drop in PHP
content management system. In my experience these programs make mysql's
problems worse or more noticeable. But, I also think config problems are
contricbuting to the problems during hight peak periods.

ken

DelphiDude wrote:

lol, msaccess. Just thinking about it a production environment is funny.

MySQL is the worst, and try as I might, I have not heard one constructive
reason for using it. I hear "regressive" reasons such as: "it's good
enough" or "It does what I need." These are not "strong" reasons, they are
passive acceptance. "It's good enough" until it isn't. "It does what I
need" until it doesn't.

For something like PostgreSQL, what is the open version of Borland's
Interbase called again? or other database, one chooses it for its technical
merits, i.e. scales under load, transactional integrity, views, functions,
stored procedures, ACID compliance, subselects, replication features, etc.

When an engineer chooses a "real" database, they can always give a technical
reason why they chose it instead of something else. When ever someone
chooses MySQL, 99.999% of the time the reason is either "It's good enough"
or "It does what I need." Which of course means, "It's good enough" until
it isn't and "Does what I need" until it doesn't.

On Fri, Mar 05, 2004 at 10:49:49PM +0000, mlw wrote:
Firebird. Notable at least as being the first RDBMS that was supported
for ESQL with tinycobol.

--
GNU/Linux revenues last quarter: $1 Billion.
micros~1 revenues last quarter: $4 Billion.
It's no longer a question of windows or GNU, it's a question of *Unix* or GNU.

ken wrote:

I don't know, and that information is not relevant to my post. You may very
well have visited the site multiple times in the past, I sure have.

Maybe so, but all the charts and graphs that I have seen, and all the
personal experience I have, says that once you get a few concurrent
operations happening on MySQL, once you have a writer or two, you loose
what ever marginal speed benefit MySQL is supposed to have.

Can MySQL have concurrent writers for the same table? Yes, but its
transactional components don't have nearly the speed of its
non-transactional default setup. PostgreSQL, with transactions and MVCC, is
close to the non-transactional MySQL in a single client scenario, and
totally kicks its ass once you get a few concurrent clients connected.