RE: one long lived connection or many short lived?

[David Brodbeck]

> -Original Message-
> From: Tommy McNeely [mailto:[EMAIL PROTECTED]

> I am building an IRC based application "bot" (using libmysql) .. that 
> will take commands from users (!mybugs, !mybugs KEY !newbugs, 
> etc), do 
> SQL queries and of course dump formatted results back to the channel. 
> Currently its setup to open a single DB connection at initialization 
> and use that connection over and over... would it be better 
> or worse to 
> have it open/close a connection for each "command" in the way 
> a web app would?

There's a tradeoff here.

One connection is much more efficient.  If you're expecting frequent
commands this is definately a concern.  However, you'll need good error
handling, so your app can reconnect when that connection drops.  Granted,
you should have this *anyway*, but with the 'lots of little connections'
scheme you get a chance to start over every time a new command is issued.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Comparing bug in 4.1.7

[David Brodbeck]

> -Original Message-
> From: Roger Baklund [mailto:[EMAIL PROTECTED]

> This definition (from the manual) is self-contradicting: 1 OR NULL 
> should evaluate to 1 because "any operand is non-zero", but it should 
> also evaluate to NULL because "any operand is NULL".
> 
> http://dev.mysql.com/doc/mysql/en/Logical_Operators.html >
> 
> With an unclear definition, it is perhaps not so strange that 
> behaviour will change with different versions of MySQL...?

I suspect the reason "1 OR NULL" returns 1 in some versions is that the OR
operator "short circuits", like in C.  A "short circuit" operator is one
that stops executing code as soon as the result is clear.  In the case of
OR, that means if you find one operand that's true, you don't have to check
the other operands.  (OR also short-circuits in Perl, which is why
statements like 'do_something() or die("It didn't work!")' work as expected.
If do_something() returns true, the die() is never executed.)

What does the SQL standard say about this, if anything?  I thought the
result of any operation on NULL was supposed to return NULL, but I could be
wrong.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: [OT] Email heaaders and threading (was Re: update MySQL)

[David Brodbeck]

> -Original Message-
> From: Michael Stassen [mailto:[EMAIL PROTECTED]

> This tells the recipient's email client that your message is 
> a reply, not a new message, despite your efforts to change the subject and

> recipients.  Many email clients use that header to decide which thread a
message 
> belongs to.  That's actually the point of the header.

I get it now. I wasn't aware of that, since every email client I've ever
seen seems to thread strictly by subject.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: update MySQL

[David Brodbeck]

> -Original Message-
> From: Jeff Smelser [mailto:[EMAIL PROTECTED]

> On Tuesday 05 October 2004 01:56 pm, David Brodbeck wrote:
> 
> > Saves having to retype the list address, or look it up.  I 
> don't see what
> > difference it makes...
> 
> Click on the email, on mine, it brings up a nice to empty 
> message with the 
> email address..

Well, that's nice...

I just don't see what difference it makes.  As far as I can see, the outcome
is identical either way...

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Dyna III electronic ignition install questions R80/7

[David Brodbeck]

> -Original Message-
> From: Brian Mansell [mailto:[EMAIL PROTECTED]

> What build of MySQL are you running? I don't believe an electronic
> ignition is included in 4.0 or 4.1.

It's slated for 5.0, to complement the turbocharger and MegaSquirt support.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: update MySQL

[David Brodbeck]

> -Original Message-
> From: Jeff Smelser [mailto:[EMAIL PROTECTED]

> Whats the deal and this list? No one can ever just hit new 
> message, they 
> always hit reply and put a new subject in..

Saves having to retype the list address, or look it up.  I don't see what
difference it makes...

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Help With MySQLdump needed!!

[David Brodbeck]

> -Original Message-
> From: Sparks [mailto:[EMAIL PROTECTED]

> I want to have an automatic backup done of my SQL databases, 
> but cant quite
> figure out how to use mysqldump to do this properly...

I use a script that I run as a nightly cron job.  Here's the one I use:

---

#!/bin/sh

DATE=`date +%Y%m%d`

cd /var/lib/mysql-backup
/usr/bin/mysqldump --user=backup --password=Password --quick --lock-tables
 --all-databases | gzip >"mysql.$DATE.gz"

cd /var/lib/mysql-backup/logs
mv update.log "update.$DATE.log"
/usr/bin/mysqladmin --user=backup --password=Password flush-logs
gzip "update.$DATE.log"

---

Obviously you'd want this to be readable only by root, since it contains
passwords.  The "backup" user will need global SELECT and RELOAD privilages.

You could easily modify this to FTP or SCP the files when they're done.  If
you don't care about the update logs you can simplify a bit and eliminate
the RELOAD privilage, but I personally find them useful to have.  (With a
backup and the update log that follows it, you can restore to any point by
restoring the backup, editing the log, then rerunning it against the
server.)

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: on big mysql .MYD files and linux's 2GB file size limit

[David Brodbeck]

> -Original Message-
> From: Eric Gunnett [mailto:[EMAIL PROTECTED]

>   Depends on what file system you are running that is a 
> ext2 limit if you move to ext3 or reiser you should be fine. 
> But then a default config on MySQL will run into a 4.29 Gig 
> limit, which you can change.

Ext2 has not had a 2 gig filesize limit for a long time.  I've created files
as large as 16 gigabytes on ext2fs with no problems.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Stop Scrolling

[David Brodbeck]

> -Original Message-
> From: Hassan Schroeder [mailto:[EMAIL PROTECTED]

> [EMAIL PROTECTED] wrote:
> 
> > When I do
> > mysql>SHOW TABLES \p less;
> > I get the error message "You have an error in your SQL syntax near 
> > 'less' in line 1"   I am using dos, not unix.
> 
> 1) that's not the example provided
> 2) \p != \P
> 3) `pager` doesn't work in Windows, anyway, per:
> 
>

Might I suggest MySQL Control Center as a more elegant way to handle this
sort of thing in Windows?
http://www.mysql.com/products/mysqlcc/

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Very Strange data corruption

[David Brodbeck]

> -Original Message-
> From: David Griffiths [mailto:[EMAIL PROTECTED]

> MySQL really should throw an exception/error rather than just quietly 
> trim your data and accept it. When your data is critical, and your 
> business depends on it, you can't have bad data quietly going 
> into the database.

The client software ought to be range-checking the data before sending it to
the database.  If the client isn't even doing that kind of minimal-effort
check, how likely is it to be checking for exceptions?

That's not to say that an error or exception is a bad idea, but MySQL may be
constrained here by what the SQL standard says to do; I'm not sure.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Gripe with MySQL

[David Brodbeck]

> -Original Message-
> From: Michael McTernan [mailto:[EMAIL PROTECTED]

> > if you don't have error checking in yourself, that's your problem
> > not mysql's problem.
> 
> True, but still doesn't help when someone inexperienced with 
> MySQL makes a
> mistake and buggers the database... again.  I don't see why 
> you object to
> making things easier to use, even if it is only activated by 
> a -n00b option
> in mysql_safe.

It also doesn't help when you have ODBC-connected software from other
vendors that does hostile things.  We've got one app here that likes to try
to insert records with a primary key of "" in one of our databases.
Unfortunatly "" and NULL aren't the same thing, so setting NOT NULL doesn't
help.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: BOM Select statement

[David Brodbeck]

> -Original Message-
> From: Wesley Baker [mailto:[EMAIL PROTECTED]

> What I want is to know how to build  a SELECT statement which will
> return all components of mynewpc and navigate down the 
> bill-of-materials
> and also return all the components for the case, the powersupply and
> even the parts on the powersupply.

This situation calls for recursion.  I don't think you can do it with just
SQL.  Maybe there's a way to do it that I'm not seeing, though.

Faced with the same situation, with Access as the front end, I ended up
writing a program in Visual Basic for Applications to do it.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: best-practices backups

[David Brodbeck]

> -Original Message-
> From: Madscientist [mailto:[EMAIL PROTECTED]

> We use this mechanism, but we do our mysqldumps from a slave 
> so the time doesn't matter.

Excellent idea.

> Interesting side effect: A GZIP of the data files is _huge_. 
> A GZIP of the 
> mysqldump is _tiny_. For our data it seems there is a lot of 
> repetition.

I think the difference is probably that the mysqldump file doesn't contain
any index data.  On some of our tables the index file is bigger than the
actual data file.  And yeah, the dump files (and update logs!) do compress
really well.  The fact that they're text and contain a lot of repeated
commands means they pack down quite small.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: best-practices backups

[David Brodbeck]

> > -Original Message-
> > From: Michael Collins [mailto:[EMAIL PROTECTED]

> > Is there any "best-practices" wisdom on what is the most preferable
> > method of backing up moderately (~10-20,000 record) MySQL 4
> > databases? A mysql dump to store records as text, the 
> format provided
> > by the BACKUP sql command, or some other method?

I think it depends on how long a backup window you can tolerate.

On our databases, I use mysqldump to dump to a text file.  The backup is
piped through gzip to reduce the size on disk.  This has the advantage of
being portable between architectures, robust, and human-readable.

I also run with update logging turned on, and turn over the logfiles after
each backup.  This way I can restore the database to any point in time by
restoring the next earliest backup, then running as much of the update log
as necessary to get to the desired point.  I use a script with the 'find'
command to weed out old backups and update logs older than a certain number
of weeks.  All of this is run nightly by cron entries.

The disadvantage of this method is that mysqldump can take a while to dump
large databases, and the tables are locked during the backup process.  If
this is a problem, you should probably investigate mysqlhotcopy.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Odd Rounding?

[David Brodbeck]

> -Original Message-
> From: Andrew Kuebler [mailto:[EMAIL PROTECTED]

> Can anyone explain why:
> 
> SELECT ROUND(.012345, 5)
> -> .01234
> 
> Why doesn't mysql round the 5 up to .01235? How do I get it 
> to round up?

Oftentimes standard practice is to round up if the digit before the 5 is
odd, and round down if it's even.  (This averages out the bias you otherwise
get with 5/4 rounding.)  To see if this is what you're seeing, try rounding
0.012335 to 5 places...I bet it'll round up to 0.01234.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Mysql listed in spamcop?

[David Brodbeck]

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]

> Somehow the mysql mailserver got listed in spamcop?

Probably the usual SpamCop GIGO problem.  If people send false reports,
systems that don't deserve it get listed.  I don't use Spamcop anymore for
that reason.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: read: Off subject regarding the mysql list!

[David Brodbeck]

> -Original Message-
> From: Bob Loeffler [mailto:[EMAIL PROTECTED]

> I have my rule in Outlook set up as:
> 
> "Apply this rule after the message arrives
> with lists.mysql.com in the message header
> move it to the MySQL and PHP folder"

I was about to suggest the same thing, except that I'm using this header to
trigger on: "List-ID: ".  Either way should work.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Mysql just stopped working on my server - any way to track do wn the problem?

[David Brodbeck]

> -Original Message-
> From: Matt Babineau [mailto:[EMAIL PROTECTED]

> I am running 4.0.15-standard on RH9. My mysql database just stopped
> working, is there a way I can log information about why it stops like
> this? the *.err was unhelpful.

Did you just install a glibc upgrade/bugfix?  That was the culprit when this
happened to me on RH7 a while back.  I'd upgraded glibc, but hadn't upgraded
mysql, which was linked against the old version and didn't like the new one.
Why up2date let me get away with that, I don't know, but it did.  Frustrated
me for *hours*.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: mysql eating up processor

[David Brodbeck]

> -Original Message-
> From: Travis Reeder [mailto:[EMAIL PROTECTED]

> Is there a way to give mysql a lower priority?

If you're on a UNIXish operating system, you could try running it with the
'nice' command. 'nice' lets you set the priority of the process you launch
with it.

Note that this only helps if it really is a CPU-bound process.  It can still
slow down the rest of the system if it's causing a lot of swapping or disk
activity.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Why does the oracle listen on this list

[David Brodbeck]

> -Original Message-
> From: Scott Haneda [mailto:[EMAIL PROTECTED]

> Pretty much every time I post to mysql list, I get one of the 
> below back,
> what is this all about and how can I stop it...

Looks like, somehow, the Internet Oracle got subscribed.  Maybe a virus
forged mail that appeared to be from it to the list subscription address?

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Best Pratices for mySQL Backups in Enterprise

[David Brodbeck]

> -Original Message-
> From: Keith C. Ivey [mailto:[EMAIL PROTECTED]

> It's true that mysqldump doesn't put the data for the index itself 
> into the dump file.  That does not automatically mean that the dump 
> files will be smaller than the .MYD, .MYI, and .frm files combined.  
> The dump file will certainly be larger that the .MYD file -- perhaps 
> much larger if you have lots of non-text columns.  The difference may 
> be greater or smaller than the size of the .MYI file.  It depends on 
> your data and your indexes.  In most cases I'd say that whatever 
> difference there is isn't enough the affect the choice of backup 
> method.

I'd agree.  Disk is cheap.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Best Pratices for mySQL Backups in Enterprise

[David Brodbeck]

> -Original Message-
> From: Keith C. Ivey [mailto:[EMAIL PROTECTED]
> Sent: Thursday, June 26, 2003 5:49 PM

> The dump file will be larger than the MyISAM data file for the 
> original table -- especially if you have many non-text columns 
> (dates, numbers, ENUM columns, etc.).  In some cases, when you have 
> large indexes, the index file will be large enough that it and the 
> data file combined will be larger than the dump file, but in some 
> cases it won't be.  I wouldn't consider the difference in size, 
> whichever way it goes, to be significant in deciding between backup 
> methods.

FWIW, if size is a problem, mysqldump files compress quite well with gzip.
(Lots of repeated text.)  The same is true of update logs, if you're keeping
those.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Anyone had a chance to try an Opteron yet?

[David Brodbeck]

> -Original Message-
> From: Lenz Grimmer [mailto:[EMAIL PROTECTED]

> Actually, you can create larger files on 32bit Linux systems 
> as well. It's
> just that the file system and the C library must have support for LFS
> (Large File Support):
> 
> http://www.suse.de/~aj/linux_lfs.html

I can confirm this.  I've been able to create files bigger than 2 gigs on
ext2fs for quite a while now.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Big Mistake Need Help, can I undo an action I did?

[David Brodbeck]

> -Original Message-
> From: Fred van Engen [mailto:[EMAIL PROTECTED]

> Sorry, you'll need a backup. If your backup is not recent 
> enough and you
> have a binary log (e.g. for replication), you can use that 
> log to restore
> the database state.

If you have a backup and the update log since the last backup, you can
restore to any point in time from that back up to present, as well.  Comes
in handy when a member of the programming staff comes up and says, "Can you
undo the 'DROP TABLE' I just did by mistake?" ;)  This only works if you've
been running with the update log enabled, of course.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Why are my sorts so slow?

[David Brodbeck]

MySQL can use indexes to sort, but not when your WHERE clause contains a
column other than the one being sorted on.

I think the MySQL manual has a good section on what 'filesort' means and
when MySQL uses it.  Basically it's selecting all the rows that meet your
WHERE clause, then running a quicksort algorithm over the results.

One thing you can do to speed up this kind of query is reducing the number
of rows it has to sort before it can select the last 10, by using a more
restrictive WHERE.  For example, I had a query very much like this that
displayed the last 50 records to be put in the database sorted by time.
Since I knew more than 50 records a day came in, I added a 'WHERE Time >'
clause to only select records from the last 24 hours.  This made a huge
difference because it was only selecting a couple hundred records instead of
a million or so before doing the sort and throwing away all but the last 50.

I don't know if that was clear or not, but what I'm trying to say is, if you
can limit the amount of data that matches the WHERE that you're going to be
throwing out anyway, do it. :)

> -Original Message-
> From: Brent Baisley [mailto:[EMAIL PROTECTED]

> Sorts don't use indexes, for the most part, only the search 
> part does. 
> Someone correct me if I'm wrong on that, I think I recall 
> reading it in 
> SQL for Smarties.

> On Friday, June 13, 2003, at 02:08 PM, Bruce Leidl wrote:
> 
> >
> > I'm having a problem with some very slow queries that spend a very 
> > long time in the 'Sorting result' state and I'm wondering how sorts 
> > are implemented in mysql and what I can do to optimize 
> these types of 
> > queries.
> >
> > The query looks something like this:
> >
> > SELECT col1,col2,col3 from table1 where col1 = 0 ORDER BY col2 DESC 
> > LIMIT 10;
> >
> > Both col1 and col2 have individual indexes (idx_1, idx_2)
> >
> > Running an EXPLAIN on the query gives me the following information:
> >
> > table:  table1
> > type:   ref
> > possible_keys:  idx_1
> > key:idx_1
> > key_len:4
> > ref:const
> > rows:   10269
> > extra:  where used; Using filesort
> >
> > I understand that this is retrieving a lot of rows which is 
> slow, but 
> > after the rows have been retrieved shouldn't the sort execute very 
> > quickly using the index on the sorted column?  I notice that the 
> > EXPLAIN makes no mention at all of the index on the column being 
> > sorted.  What is a filesort and what are the different ways 
> that mysql 
> > can sort a result set?

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

Other front ends (Was: RE: I need advice. MSAccess frontend to PH P/MYSQL. ANyone done this?)

[David Brodbeck]

> -Original Message-
> From: MyLists [mailto:[EMAIL PROTECTED]

> The main thing you should realize right off the bat is that 
> Access gives you
> a lot of functionality with buttons and other controls and 
> they are quite
> easy to control with macros and/or VBA. However, depending on 
> the complexity
> of your forms, most of this functionality will be a challenge 
> to replicate
> on the web, regardless of which language you choose. 

Are there any open-source database front-ends that have similar
functionality to MS Access?  We're currently using Access with tables linked
to MySQL, but find it leaves something to be desired when it comes to
stability and license flexibility.  We'd really like to move to something
open-source.  PHP doesn't really fit the bill because a web-based interface
isn't ideal for our application -- we prefer the more interactive feel of
Access's forms.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Is port forwarding over SSH an OK way to do it?

[David Brodbeck]

> -Original Message-
> From: Peter Rønning [mailto:[EMAIL PROTECTED]

> I'm planning to set up an MySQL database on a server, and 
> connect to it
> over the internet (no connection to a website or anything, just a
> private db server). Just wondering what is the recomended way 
> to do it,
> port forwarding over SSH, or using SSL. I'm leaning towards using SSH,
> since I have already gotten that to work and it provides a secure
> channel for the data transfer (which is all I need, really). 

It doesn't make any difference from MySQL's perspective whether you use SSL
or SSH.  I'd use whichever you're more comfortable with and find suits your
needs better.  SSH is a little more flexible when it comes to authentication
method.  SSL is more widely supported in software libraries if you plan on
writing a custom client.

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: How to secure a MySQL database from people with physical acce ss

[David Brodbeck]

> -Original Message-
> From: mos [mailto:[EMAIL PROTECTED]

> I could encrypt certain table fields, but this will make 
> writing the front 
> end a pain because all SQL statements will now need to be 
> changed any time 
> a new column is encrypted.

It also won't help you any, because the software will have to contain
everything needed to do the decryption.  Unless you can somehow prevent a
hypothetical attacker from getting this software, your encryption is only
going to keep a casual attacker out.  All he has to do is decompile the
software enough to figure out your encryption routine.

Generally there's very little you can do to protect data from someone with
physical access to the machine -- unless you can keep it in encrypted form,
and only decrypt it elsewhere, so that the decryption key never passes
through the vulnerable machine.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Could we make this a web discussion forum?

[David Brodbeck]

> -Original Message-
> From: Seth Brundle [mailto:[EMAIL PROTECTED]

> > I don't have to wait for some remote, overloaded server
> > to respond.
> 
> I dont understand this one at all.

Every web forum I've used has been slow.  I don't want to have to wait 10-20
seconds to pull up each message when I'm browsing, when in an email list my
client can pull them up instantly.  It's frustrating and a waste of my time.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Could we make this a web discussion forum?

[David Brodbeck]

> -Original Message-
> From: Seth Brundle [mailto:[EMAIL PROTECTED]

> This makes no sense as message I post to a Yahoo! Group get emailed to
> opt-in members and appears on the group within seconds, while 
> I may not
> receive something I posted to the MySQL list for minutes or 
> sometimes hours.

It's not the turn-around time I'm referring to.  I can go do something else
during that.  It's the time to pull up each message.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Could we make this a web discussion forum?

[David Brodbeck]

> -Original Message-
> From: Seth Brundle [mailto:[EMAIL PROTECTED]

> I've worked with MySQL for about 7 years 
> and have never
> found it convenient.

Fair enough.  But I've worked with probably half a dozen web boards and have
yet to see one I liked better than an email list.  All the ones I've tried
have been awkward and much more time consuming to use.  Anything delivered
through HTTP that plays at being realtime and two-way ends up being slow and
painful.  It's just not what the protocol was designed for.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Could we make this a web discussion forum?

[David Brodbeck]

> -Original Message-
> From: Seth Brundle [mailto:[EMAIL PROTECTED]

> H...seems like some Yahoo! employee  has never used  
> Yahoo! Groups ;)

I know a lot of people who refuse to use Yahoo Groups because of Yahoo's
very open and ever-changing privacy policy.

Most of the other combination web discussion board/mailing list systems I've
seen are not very good at doing mail.  They're web chat boards with mail
notification tacked on.

> If every package I used involved a mailing list for discussion, I would
need to track 
> about 40 of them.

How will web boards help?  It's easier for me to glance through a stack of
email folders (sorted by my mail rules) than it is for me to visit a bunch
of web sites every day.  Between logging in to each one individually,
waiting for sluggish page loads, and digging through all the nested
discussion folders it's just too time-consuming.  It's much better for me to
have my mail client collect everything in one spot for me.

> Email is ok for thread tracking if subject integrity is 
> maintained, but most
> email clients cant reliably collapse and expand threads.

Microsoft Outlook seems to do it well.  Are you telling me no one but
Microsoft has been able to get this right?

> Sorry, I didnt realize that I didnt need to subscribe to post - but again,
> this is one of those rules for each mailing list that you must remember,
> which is a PIA.

For me the situation is worse with web forums, because every one of them
works differently.  It's confusing.  Plus I have to remember a username and
password for each one, and log on individually, instead of logging on once
to my mail account.  I avoid web forums for these reasons.  They're just a
lot more painful to use than a mailing list.  

They're also almost always really sluggish and full of graphical animated
.signatures and brightly colored smiley-face cruft.  Just not worth the
trouble, unless you're a 14-year-old script kiddie who still thinks that
stuff is fun. ;)

I think if the MySQL list went to a web-only discussion board I'd probably
stop paying attention to it.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: Could we make this a web discussion forum?

[David Brodbeck]

> -Original Message-
> From: Jeremy Zawodny [mailto:[EMAIL PROTECTED]

> I really hate on-line forums.  They're difficult to track because I
> must remember visit them daily.

Agreed.  I don't need yet another web page to keep track of.  I prefer to
let my mail reader thread up the messages on this list, then I browse
through every now and then and delete everything with a subject that doesn't
look interesting.  It's much more convenient than a web forum, and much
faster too, because I don't have to wait for some remote, overloaded server
to respond.

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

RE: ODBC connection from Windows app to Linux backend?

[David Brodbeck]

Would MyODBC do what you want?  I'm using it to connect MS Access to MySQL,
and also to connect a Borland Delphi application to a MySQL database.  It
works pretty well.  Consult the FAQ for known glitches with specific
applications.

> -Original Message-
> From: LaBranche, Kevin [mailto:[EMAIL PROTECTED]
> Sent: Wednesday, March 26, 2003 5:02 PM
> To: '[EMAIL PROTECTED]'
> Subject: ODBC connection from Windows app to Linux backend?
> 
> 
> How would I connect my Windows app to a linux MySQL DB.
> 
> Would I just modify the DSN I am using to point to the correct server?
> 
> Thanks,
> Kevin
>  
> -Original Message-
> From: LaBranche, Kevin [mailto:[EMAIL PROTECTED] 
> Sent: Wednesday, March 26, 2003 12:18 PM
> To: '[EMAIL PROTECTED]'
> Subject: Query not returning what I expect
> 
> 
> The following query returns 0 for sCode instead of a string like
> "XXX:XX".
> 
> I am modifying an existing app from SQL Server 7 so that it 
> can also run
> with a MySQL backend 
> 
> The existing query in SQL Server 7 works fine
> SELECT sDAMIONCode, (sORI + ':' + sDamionCode) as sCode FROM tblORI 
> 
> Any hints would be appreciated. 
> 
> Kevin
> 

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

Need some help accessing unsigned columns from Delphi

[David Brodbeck]

I'm trying to write a Delphi program that accesses a MySQL database via
MyODBC.  It's going pretty well except for a sticky problem I'm having with
an unsigned integer column.

I have a column named 'IP', type unsigned integer, that will hold IP
addresses.  However, Delphi seems to assume this column is a signed integer
column and limit the range accordingly.  At least that's my best guess for
what's happening.  For example, this code fragment:

IP := $;
Str( IP, IPstring );
tblDigital.FieldByName('IP').AsString := IPstring;

throws an exception saying that 4294967295 is not a valid value for this
field.

Do I have any options other than giving up and using a signed BIGINT field?

--

David Brodbeck, System Administrator
InterClean Equipment, Inc.
Ypsilanti, Michigan
[EMAIL PROTECTED]
(734) 975-2967 x221



-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: timing

[David Brodbeck]

Have you checked for network problems?  You might try FTPing a file to and
from the production server, or something similar, just to rule this out.

I've had slowness problems with various apps that were driving me *nuts*
until I found a network issue causing them.  In a few cases I had a duplex
mismatch -- an ethernet card set to full duplex connected to a half-duplex
hub/switch, or vice versa.  This results in a connection that appears fine
when you ping and works great under light traffic, but bogs down with
massive numbers of collisions as soon as you do anything that involves
moving lots of data.  The collisions may not show up in the statistics since
the end that's set to full duplex won't notice them.  Letting the card
auto-detect the duplex setting does *not* always work right...

This is a subtle, easy trap to fall into.  Easy enough that I've done it
multiple times. ;)

> -Original Message-
> From: Mary Stickney [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, January 30, 2003 9:08 AM
> To: [EMAIL PROTECTED]
> Subject: timing
> 
> 
> 
> I have 2 Crystal Reports 8.5 (for the record I didn't make 
> these reports)
> that draw straight from the MYSQL
> database. They use only 1 table. I am calling them to print 
> from VB 6.0
> Theses reports have lots of complex calculations in them, 
> counts , sums and
> groups.
> 
> My test server has the same code , same database , same indexes , same
> amount of memory as the Production server,
> My test server has MYSQL 3.23.51 and MYODBC 3.51 , and I have Crystal
> installed on my test server.
> 
> The Production server has MYSQL 3.23 and an older version of 
> MYODBC then I
> have.
> No Crystal Reports installed. And the server is a faster 
> machine then mine.
> 
> 
> They run in 8 minutes on my test server , about 4 minutes 
> each. Printing to
> a PDF driver.
> 
> On the Production Server they take HOURS to run and print.
> 
> Any ideas...
> 
> 
> 
> 
> Mary Stickney
> TAG-TMI
> Data Warehouse Programmer
> 402-474-7612 x 3099
> [EMAIL PROTECTED]
> 
> 
> -
> Before posting, please check:
>http://www.mysql.com/manual.php   (the manual)
>http://lists.mysql.com/   (the list archive)
> 
> To request this thread, e-mail <[EMAIL PROTECTED]>
> To unsubscribe, e-mail 
> <[EMAIL PROTECTED]>
> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
> 

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: ReiserFS vs Ext3

[David Brodbeck]

> -Original Message-
> From: Michael T. Babcock [mailto:[EMAIL PROTECTED]]

> I would E-mail the reiserfs list and ask for whether ReiserFS 
> would be 
> better for a MySQL-only partition than ext3 and see what answers you 
> get.  In the worst case, you'll have some extreme "resierfs is always 
> better" responses that don't help you decide.  In the best 
> (and what I 
> think will happen), someone will tell you the differences in 
> CPU usage 
> or fast file access speeds or some such value that will mean 
> something, 
> or even that it is not actually going to make a difference.

If Nicholas does this, I hope he'll report back here with what he finds out.
I'm facing the same choice very soon, for a database that will eventually
have millions of entries (but each individual entry very small.)  I'm trying
to decide whether to go with reiserfs or ext3.  ext3's my current favorite,
but only because it's more of a known, stable quantity.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: myODBC, if left alone, can lock up an Access table

[David Brodbeck]

> -Original Message-
> From: Steve Quezadas [mailto:[EMAIL PROTECTED]]

> I notice that sometimes when I establish a connection with 
> myODBC with 
> Microsoft Access, and send it occasional sql statements it 
> works fine at 
> first. But if you let it sit for, say, 30 minutes and comes back, the 
> Access table locks up. I am not sure if this is a problem with myODBC 
> (which I would report to this mailling list) or a problem with Access 
> itself (thus I would have to report it to the Access 
> newsgroup). Anyways, 
> this problem is kinda annoying because the only way I can get 
> out of it 
> is if I force-quit Access.

I haven't seen this problem.  The only time I've had to force-quit Access
was if I changed a table definition or restarted MySQL while Access was open
-- this tends to put it in an infinite loop.  (Great error handling,
Microsoft!)

Maybe someone can help you out more of you provide more details of your
setup?  MySQL, MyODBC, and Access versions, operating system of the clients,
etc.  Linked tables, or direct access via VBA?  Anything else unusual?

FWIW, we're using Windows NT 4.0, Access 2000, and MyODBC 2.50.39.  We're
using linked tables.  We haven't had any problems that weren't covered by
the MyODBC FAQ.  I highly recommend following their recommendation of adding
a TIMESTAMP column to each table, it fixed the flakiness we were having with
disappearing records and locked tables.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: RE: Problem to access MySQL via MySQLFront

[David Brodbeck]

> Stupid filter fodder: sql,query,queries,smallint

> > -Original Message-
> > From: Dean Householder [mailto:[EMAIL PROTECTED]]
> 
> > The problem is with the new glibc-2.2.5-40 package released 
> > on October 3,
> > 2002. Apparently its a known issue, and people are aware of 
> > it. The best
> > solution I've seen so far is to role back to a previous 
> > version of this
> > package, and its dependencies.
> 
> There was a subsequent glibc update that fixed this issue, if 
> I remember
> right.  I upgraded again, to the very latest one RedHat has 
> released, and it
> solved the problem.
> 

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: What, if anything, is wrong with UNIX Epoch time stamps? [Was: R E: TimeStamp in MySQL reqd NULL]

[David Brodbeck]

> -Original Message-
> From: Michael T. Babcock [mailto:[EMAIL PROTECTED]]

> Csongor Fagyal wrote:
> 
> > Oh and one more thing I am not really sure of: sometimes 
> defining the 
> > day as 3600*24 or the year as 3600*24*365 is not the best 
> idea... just 
> > think about leap years. How do you handle that? And there are some 
> > more "artifacts" in the Gregorian calendar, too...
> 
> 
> But MySQL doesn't guarantee correctness in time values in the first 
> place.  You can still insert "2002-02-31" as a date if you like:

True.  But the example of 3600*24*365 is obviously wrong.  It should at a
minimum be 3600*24*365.25, though that might cause some odd artifacts.  The
best idea would be to put logic in the code to implement the leap year
rules.  (Which are more complicated than you learned in elementary school!
;)

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: How long is my piece of string?

[David Brodbeck]

> -Original Message-
> From: Mike Wexler [mailto:[EMAIL PROTECTED]]

> The best answer is, "try it and see." If your tables are 
> small. Adding 
> and removing indices is very quick. And testing queries is 
> also very quick.

Enabling the slow query log, and enabling log-long-format can give some
hints by logging all queries that don't use indexes.  Then you can feed
those queries to the EXPLAIN command and figure out why.  I dramatically
sped up several queries on my database this way.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: Is it possible to backup a corrupt database without being

[David Brodbeck]

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]

> I can FTP and telnet in to my site as "admin" or a user, but 
> all the files in 
> /var are owned by root so cannot be deleted, with the 
> exception of several 
> mysql files owned by mysql, which I also cannot move or 
> delete :(  If I could 
> extract the info in the database I could delete that 
> remotely, then restore it 
> after I get root access back again, but as it is now marked 
> corrupt I can't 
> seem to do anything with it.  Is there any way of fixing a 
> database with zero 
> room on the partition it is in?

Just a thought: FTP the database files to another machine.
Drop the affected databases to delete the files
 and free up enough disk space to log in.
Log in as root and free up more disk space.
FTP the database files back to the server.
Repair the database.

My apologies if you've already thought of this.  I don't have any experience
with the raq3, so maybe this is unworkable.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: Can MySQL handle 120 million records?

[David Brodbeck]

> -Original Message-
> From: Joe Stump [mailto:[EMAIL PROTECTED]]

> We hit the 2gb file limit in Linux (NOT a MySQL problem) and 
> moved to Solaris without incident.

This appears to have been largely fixed in Linux, too, if you use a recent
kernel and glibc.  I recently tried creating a 3 gigabyte file on one of my
ext2fs partitions and it worked fine.  I'm not sure what the new limit is.
The system is RedHat 7.0 with all the current bugfix updates (including
glibc 2.2.4), plus a custom-compiled 2.4.17 kernel.

This has been discussed a lot on the samba mailing list.  It seems to still
be a problem with some distributions and not with others, and no one's quite
sure what the deciding factor is.  It's easy enough to test by dd'ing a few
billion bytes from /dev/zero into a file, though.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: access->mysql

[David Brodbeck]

I used MySQLFront.  It wasn't entirely automatic, but it did a reasonable
job.  There are some gotchas.  

Take your time, and be sure to read the notes for using Access and MyODBC
with MySQL, if that's how you're going to do it -- some column types are not
recommended.  You'll want to add a TIMESTAMP column to each table, as well,
after importing.  Some of MySQLFront's choices for column types aren't the
best -- you should check them by hand before doing the import.  In
particular it doesn't handle currency fields properly unless you manually
set the type.  (It defaults to Integer, and truncates the decimal parts.)

If you have any table names with spaces, fix those before you try to import.
MySQLFront will happily create them, but you'll find them impossible to
actually use.

You'll have to do some hand editing of your Access queries.  Access doesn't
support relationships between ODBC data source tables, so you'll need to
manually set up the joins in each query.  Keep in mind that Access will not
enforce data integrity on ODBC tables.  Also remember that Access is not
case-sensitive, but MySQL is!

You should try hard to get the tables correct in MySQL before you deploy the
database to users, because after that changing them gets more difficult.  If
a table definition is changed, you have to go back and delete the link to it
in Access, then recreate it.  If you don't do this, Access gets confused and
goes into an endless loop.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: [OT] LEGAL information about MySQL.

[David Brodbeck]

> -Original Message-
> From: Pae Choi [mailto:[EMAIL PROTECTED]]

> Seems like all against I see are coming from MySQL team.
> I know MySQL is stealing a lot of code from the public
> domain, inclduing the SSL part. How funny!

By definition you can't "steal" anything from the public domain.  It's free
for the taking!

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: backing up mySQL database ?

[David Brodbeck]

> -Original Message-
> From: Neil Tompkins [mailto:[EMAIL PROTECTED]]

> Can anyone recommend a software app. I could buy to be able 
> to dump data and 
> table structures to a text file.  Ideally as mentioned before 
> I need to run 
> this on a hourly basis.

Backing up the entire database on an hourly basis could be time- and
processor-intensive, if the database gets large.  I'd recommend backing up
the full database at longer intervals (say, nightly) with mysqldump and then
using the update log to fill in any gaps.  If getting the backups onto a
different machine is important you can always just copy the log.  A backup
plus all the update log entries since the backup was taken is sufficient to
recover the database.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

RE: MySQL hangs on remote connections

[David Brodbeck]

Never mind, I found the answer to my own question when I got the archive
search engine to respond.  (It's *very* slow from here, for some reason --
about 10 minutes per query.)  I installed a newer version of the glibc fix
and all is well now.

Sorry for wasting everyone's time.

> -Original Message-
> From: David Brodbeck [mailto:[EMAIL PROTECTED]]
> Sent: Wednesday, November 27, 2002 9:09 AM
> To: '[EMAIL PROTECTED]'
> Subject: MySQL hangs on remote connections
> 
> 
> I'm having a rather bizarre problem...at least, it seems 
> bizarre to me.
> 
> We have a small MySQL database that was running great for 
> about a month.
> Last night, I rebooted the system it runs on.  Now I can 
> connect locally,
> but remote connections fail.

-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

MySQL hangs on remote connections

[David Brodbeck]

I'm having a rather bizarre problem...at least, it seems bizarre to me.

We have a small MySQL database that was running great for about a month.
Last night, I rebooted the system it runs on.  Now I can connect locally,
but remote connections fail.  The server disconnects during the query and I
get the following in the error log:

Number of processes running now: 1
mysqld process hanging, pid 17308 - killed
021127 08:58:51  mysqld restarted
/usr/libexec/mysqld: ready for connections

The MySQL version is 3.23.32, and the OS is RedHat Linux 7.0.  The only
change I can think of that's been made is the RedHat glibc update to fix the
name resolver security hole was done, and mysql wasn't restarted until now,
so it would have been running on the old glibc until the reboot.  I checked
all the tables with myisamchk and they're okay.

Any ideas?

--

David Brodbeck, System Administrator
InterClean Equipment, Inc.
Ypsilanti, Michigan
[EMAIL PROTECTED]
(734) 975-2967 x221



-
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/   (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php