MySQL Bug #11472: Triggers not executed following foreign key updates/deletes

https://bugs.mysql.com/bug.php?id=11472

by gyosko • 11 years ago

102 54 11 years ago

bugs.mysql.com

reisub• 11 years ago

Does anyone know if this is fixed in MariaDB?

AnkhMorporkian• 11 years ago

No, it is not.

https://mariadb.com/kb/en/mariadb/trigger-limitations/

51Cards• 11 years ago

Probably because they don't want to break their compatibility with mySQL, a decision of theirs I highly appreciate. Being able to drop Maria in for mySQL without having to rework a lot of things has been a huge time saver.

_lflx• 11 years ago

See you all in another 10 years.

pilif• 11 years ago

if you rely on triggers for data integrity, then I guess this is one more case where MySQL is letting you down and allowing your data to get corrupted unless you're really careful.

While it's not always good design to use triggers for this, sometimes, it's a valid reason to use them for integrity checking or enforcing. Having `on delete` triggers not run for some delete's is violating the principle of least surprise.

When feature break this way, people start to distrust them and best practices get adopted that discourage using them, killing the features all together.

Using a database that does not have this misfeature, both triggers and foreign keys are perfectly safe to use, but because the one database that is the most widely used has issues like the one described here, you often hear the recommendation to not use triggers or even foreign keys.

Yes, you can potentially move the logic normally contained in either of them into the application, but as you deal with more concurrency and as you start accumulating bugs, sooner or later your data will be messy and you will need to add cleanup-scripts, or, heaven forbid, work around messy data in your application logic (don't. do. that. it's the path to the dark side of unmaintainability).

Or, of course, use a database system whose features work correctly and the principle of least surprise is in effect.

PostgreSQL is one of these, by the way.

kailuowang• 11 years ago

Yes, you can potentially move the logic normally contained in either of them into the application, but as you deal with more concurrency and as you start accumulating bugs, sooner or later your data will be messy and you will need to add cleanup-scripts

I agree with most of what you said, but in many many use cases data integrity can be maintained in application code (a statically typed system helps), it simply can't be true that all those applications running on NoSql database will have messy data and/or cleanup-scripts.

spotman10• 11 years ago

furthermore, foreign keys are not performant at mass scale on both pg or mysql or any traditional RDBMS. if your trying to do several thousand ops per second, the vector of index contention greatly goes up if there is FKs littered all over your schema.

nosql or sql large platforms almost always have some amount of work always ongoing to clean, shuttle, and maintain data integrity.

that doesn't make me not a fan of them; for lower transaction rate applications ( which most probably are ), they are a small time investment to have some sanity checking. but it's a right tool for the job thing.

this is yet another reason why people dislike FKs in best practice stuff.

Edit: Hi down-voters. Curious to hear why I'm being down voted and possibly be offer a rebuttal?

SFjulie13• 11 years ago

Apple uses MVC with sqlite for the contacts. The view is cocoa, but every events int the GUI are calling triggers in sqlite.

It makes apple's application simple stupid and efficient.

And they have triggers on foreign keys... with sqlite.

...

It is not postgres in one of it, it is much more mysql is not a correct rdbms. None of the other databases have this flaw.

Mysql is to RDBMS what mongo is to NoSQL, nowhere short of achieving anything that matters correctly.

EDIT: or better: Mysql is a realistic in the Hollywood way implementation of a RDBMS

spotman10• 11 years ago

mysql is good for very high transaction rate platforms, that mostly rely on simple features. for this, it generally outperforms most of its competition in the traditional RDBMS space, and why companies like twitter and facebook use it. mysql likes simple, well designed queries.

postgres usually outperforms mysql for more complex operations, but is not as fast as mysql for simple ones.

sqlite is really only appropriate for things like mobile or desktop applications or things with generally low concurrency requirements.

i would be careful throwing the baby out with the bath water in your dismissal of mysql, but thats just my opinion:)

can you shed some more detail on that apple uses triggers with sqlite? that is interesting to me, and would like to learn more.

icelancer• 11 years ago

Classic:

[30 Jun 2005 19:04] Dmitry Lenev We will fix this in 5.1

walrus• 11 years ago

I don't understand some of the people commenting on the bug report. If they really want it fixed, then they should fix it. Making rude comments isn't going to make it go any faster.

nightski• 11 years ago

My guess is MySQL has quite a few users who are not technically capable of fixing a deep and involved bug in the heart of a database engine.

protomyth• 11 years ago

and I would imagine some of the people filing bugs are people who are paying for support given MySQL and, later, Oracle's business model

lucb1e2• 11 years ago

As someone who had issues with a 4+ year old bug in FileZilla, can confirm. First time I developed anything serious in C++ and it took me only two evenings to develop a patch. Then some back and forth on the bugtracker, but it made it in soon after.

(For those interested, it's the ability to create a new, empty file on the server.)

breakingcups• 11 years ago

Surely the two issues aren't of the same complexity.

infamouscow• 11 years ago

Exactly.

One of the major reasons to pick open-source software over a proprietary software is you can fix bugs that the vendor or developers won't.

Twirrim• 11 years ago

It would be really awesome if MySQL crew could spend a few months (half a year?) working on paper-cuts:

https://bugs.mysql.com/search.php?search_for=&status=Active&...

That's all the active bugs against MySQL. Note that this 10 year old bug isn't the oldest.

This is the oldest, from March 2003: https://bugs.mysql.com/bug.php?id=199 Note someone submitted a fix for that over a year ago (and verified under the OCA in November) but it still hasn't shipped. This is a bug that should never have taken 11 years to fix.

https://bugs.mysql.com/bug.php?id=3052 This one is great. ROLLBACK in a stored procedure doesn't close the cursor.

MySQL, for all its strengths, often feels like there is a bunch of typical geeks running the show. Lots of focus on the 'sexy' new features, not so much focus on keeping the lights on.

morgo2• 11 years ago

There has been a lot of focus on this for MySQL 5.7. Just take a look at the list of deprecations and changes: http://mysqlserverteam.com/removal-and-deprecation-in-mysql-...

We are currently looking into bug 199. The patch itself is probably too large to backport into a GA release.

Nemcue• 11 years ago

As if suits ever wanted anything else but new sexy features.

cremno• 11 years ago

In a few days #20786 will have its 9th birthday.

https://www.youtube.com/watch?v=oAiVsbXVP6k

https://bugs.mysql.com/bug.php?id=20786

morgo2• 11 years ago

This is caused by a layering problem in MySQL itself:

- Triggers are at the SQL Layer

- Foreign keys are implemented natively by InnoDB at the storage engine layer

It is not as easy to fix as it sounds.

kstrauser• 11 years ago

It's probably not, but I'd rather not have the feature at all than think that it's there and working but actually not doing what I've asked it to. The current short-term fix would be to make it raise an error when you try to create such a trigger, so at least no one would be using it for now.

riffraff• 11 years ago

Firefox's issue #106400 (osx keychain integration) will turn 14 this year, AppEngine's #3091 (servlet 3.0 support) has already turned 5.

I guess every project as a few "we'll fix it some day" tickets.

jessaustin• 11 years ago

Somehow those seem less important. Maybe it's because I cut my teeth on Sybase and Oracle rather than noSQL, but if triggers don't work, I would really hesitate to call that a RDBMS.

needusername2• 11 years ago

I don't think it's fair to compare MySQL to GAE/J. MySQL is actively developed by a company, marked and sold with support.

xyby• 11 years ago

The question is if one should "fix" this at all. If you change the behavior from "foreign keys do not activate triggers" to "foreign keys activate triggers", everyone who uses triggers will have to audit their applications for potential problems. And for large applications, that can be a lot of work.

MariaDB documents "foreign keys do not activate triggers" as the standard behavior:

https://mariadb.com/kb/en/mariadb/trigger-limitations/

Personally, I am grateful for software that changes as seldom as possible. I don't want to spend time on "updating" my application because something down in the stack changed.

LoSboccacc• 11 years ago

Well what does SQL standard say about that?

Application can just not update, so there is no need for windows level of compatibility management

Personally never jumped on the mysql train and can say I'm really glad for that.

zzzeek• 11 years ago

triggers...bah. how about CHECK constraints? (http://stackoverflow.com/a/2115641/34549) The irony is that the workaround for CHECK constraints being silently ignored is to use a trigger.

mianos• 11 years ago

Mysql has its uses. Keeping a list of people you don't know on a social network is a great use case. Recording cat video URLS is another. Keeping track of money is not a good use case for mysql. Simple as that. My favourite is this one is the correlated subquery bad query plan selection: http://bugs.mysql.com/bug.php?id=9090 "It's not a bug, it's a feature". You can always find a way to work around this stuff, or you can use postgres or another db that considers thi stuff important.

MySQL Bug #11472: Triggers not executed following foreign key updates/deletes

Related Stories