| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Comments:
<0> zardinuk: I *do it in MyISAM. A single statement is atomic anyway
<1> with innodb, you really don't have to worry about things getting screwed up
<2> GAH!
<1> inviso: no it isn't
<0> zardinuk: sure it is. You can't select and then issue an update, but there are certainly ways to get it done
<0> Same sort of thing applies to stock counts for inventory
<1> inviso: only manual locking
<3> hi people
<3> where can i find mysql icons?
<1> and that isn't a single statement
<4> anyone know how to fix my broken slave?
<0> zardinuk: yeah, not necessary to use more than one statement typically :)
<5> is it possible to find the value of a primary key auto_increment field for a row you've just inserted into a table?
<0> !m fred87 last_insert_id
<6> fred87: (Information Functions) : http://dev.mysql.com/doc/mysql/en/Information_functions.html
<5> thanks
<2> So you mean that "transactions" ensure that WHILE I'm executing a SELECT query, for example, and somebody does something to the data I'm fetching, these queries will know about each other and "wait"/catch up with each other?
<2> Sounds like a hell to code...
<1> use wikipedia to look up MVCC
<1> not too complicated
<0> KimmoA: yup, thus why it's slower. If your app needs that kind of guarentee, use transactions
<2> This should only be NEEDED for stuff where money is involved, yes?
<1> better said "if your tables are read only, you shouldn't bother with concurrency"
<2> Is that a joke?
<7> will providing mysql with an very large insert statement (thousands of rows) cause it to disconnect the client?
<2> Or do you mean that the entries are "read only"?
<1> KimmoA: if your not using innodb, you have to write extra code to account for deadlocks and what not
<1> not deadlocks, but you have to implement your own concurrency and locking
<8> or just use artificial locks
<8> that time out
<7> getting the error on php: the mysql server has gone away
<2> zardinuk: ***uming it's a critical multi-billion dollar bank table, right? :)
<1> if you ever write to a table that is read and vise versa, you should use the built-in concurrency (innodb), if allyou ever do is write or read, not both, then you don't need to bother
<2> Well... you mean write to existing data, right?
<1> KimmoA: you give me an example and I'll tell you how it'll break
<2> In other words: UPDATE.
<2> OK. The table I have is constantly SELECTed from, and sometimes INSERTed to.
<0> zardinuk: update inventory set stock_count = stock_count - 3 where id=24; What's wrong with that?
<2> Not updated unless I need to fix something manually.
<2> It's also DELETEd.
<9> hello everyone
<10> hello one
<1> inviso: one user runs the exact same query at the exact same time, it's like only one ran
<2> inviso: Quote 24.
<1> inviso: err, two users ran the same query, rather
<2> Are you speaking to the right person?
<0> zardinuk: not true.
<9> I have a question concerning merging specialized tables
<1> KimmoA: what is selected from it
<2> zardinuk: The application?
<1> inviso: yes it's true, explain yourself
<0> zardinuk: the statement is atomic. Both will decrement by 3
<1> KimmoA: yeah, what are you selecting?
<2> zardinuk: Let's just get things nice and sparkling clean: This "transactions ****up" can only happen on non-transactions tables that do UPDATE queries, right?
<9> I have around 44 tables and most of them constitute a arborescence of specialized tables, the "father" one represents properties
<1> inviso: in order to decrement by 3, you have to select the initial value, then decrement it, then write it back... you see where the error creeps in?
<0> zardinuk: that's not how execution happens. It sets count = (count - 3)
<1> right...
<2> (I thought he aked for syntax help.)
<1> and that's broken down into 3 steps
<2> s/aked/asked/
<0> zardinuk: nope, it's atomic
<1> inviso: only if you're using innodb, you get it?
<1> it's not atomic with isam
<11> Is there any sane reason why this can happen: http://www.morbo.org/blog/2006/04/14/its-2-2/ ?
<9> I tried using inner joins to select one branch of the specialization and then a left join of two of those individual branches, but I am stuck with the syntax
<1> http://www.dbazine.com/oracle/or-articles/kyte1/
<1> there you go inviso
<0> zardinuk: http://dev.mysql.com/doc/refman/5.0/en/ansi-diff-transactions.html Please read under the section """You can also use functions to update records in a single operation"""
<0> zardinuk: that's an oracle article
<1> oracle and mysql use the same concurrency mechanism
<0> and the lost updates part is crap. That's not how atomic statements are executed. If I issued a select and then an update, yes. But, not in one statement
<0> zardinuk: just read the doc I posted you :(
<1> inviso: you don't understand, one statement is one transaction
<0> """You can also use functions to update records in a single operation. You can get a very efficient application by using the following techniques:"""
<0> """Modify columns relative to their current value."""
<0> Is that not clear?
<1> you could do select @@var for update; @@var+=3; update @@var;
<0> that would give you concurrency problems
<1> inviso: that article is talking about "statements" as it relates to InnoDB
<1> with ISAM, a statement is vulnerable to lost updates
<1> with innodb, one statement is one transaction, (auto-commit)
<0> zardinuk: dude.... Pause your mouth and read that part of the article I linked. They explicitly give an example like I gave you "UPDATE tablename SET pay_back=pay_back+125;" and explain that it is atomic.
<1> ACID isn't only talking about begin and commit, it's practically everything you do with the db
<12> zardinuk: ISAM? In year 2006?
<0> Any mods around?
<1> inviso: yes, it is ACID compliant IF YOU'RE USING INNODB, you get that?
<0> zardinuk: you're wrong.
<1> haha, you're wrong dude, what makes you think you know what you're talking about?
<0> please read the article
<1> I know exactly what you're trying to say, and you don't realize the difference between a transaction and a statement
<12> zardinuk: You better read first :)
<0> zardinuk: is UPDATE tablename SET pay_back=pay_back+125; atomic across MyISAM?
<1> no it's not
<0> zardinuk: The article states that it is. Read it.
<12> zardinuk: Can you prove it?
<1> inviso: i'm reading it, I don't see where it says it's ATOMIC
<12> zardinuk: I know the answer. I wonder if You know it or you just talk because you heard someone else saying it
<1> besides, I know these things, I've been studying database design for years
<0> zardinuk: """This is very efficient and works even if another client has changed the values in the pay_back or money_owed_to_us columns."""
<1> k inviso, genius, now read the first sentence of that article
<0> Following are some techniques for working with non-transactional tables:
<1> and report back to me
<13> Does anyone here know of a web based application I can install to enable users to view mysql databases?
<0> It's under that section :(
<14> The other non-transactional storage engines in MySQL Server (such as MyISAM) follow a different paradigm for data integrity called atomic operations.
<0> The part about non-transactional tables, IE MyISAM
<13> Kind of like a database client on the web.
<0> moj0rising: phpmyadmin
<0> !tell moj0rising about gui
<6> Alrighty.
<13> Thanks! I like and use phpmyadmin..
<1> inviso: have you accepted that I'm right yet?
<9> anyone here could explain me how to merge multiple tables?
<13> I amlooking for someting like...
<1> You can generally code around the need for row-level locking. Some situations really do need it, and InnoDB tables support row-level locking. Otherwise, with MyISAM tables, you can use a flag column in the table and do something like the following:
<12> moj0rising: Many people like phpmyadmin
<10> inviso, What?
<13> ..AdoDB on the web.
<2> This feels like TCP VS UDP.
<13> Like the openoffice DB application, only web based.
<2> "UDP can do what TCP provides but you have to code extra"...
<12> zardinuk: You can use LOCK TABLES instead of flag, but that does not mean you know the answer of the question: "Is UPDATE MyISAM_table ... atomic or not?"
<15> UDP: the data transmision for real men
<15> inviso: how'd you get ops?
<0> cj: You'd be surprised :)
<2> cj.com ****s.
<1> salle: it's atomic with innodb, not with isam
<2> They want me to LOGIN to get my p***word back...
<0> !kick zardinuk Please learn to read
<12> zardinuk: In some cases it is guaranteed to eb atomic while in other cases it is not. It's quite obvious when by the way :)
<12> inviso: isam?
<13> Does anyone know where I could get an application like that (above)?
<0> awww, he doesn't love me :)
<12> inviso: Who uses ISAM nowadays?
<5> inviso: nice :)
<1> inviso: why are you trying to get me kicked? I'm trying to help you understand something here
<0> zardinuk: Please read that article. All the way through.
<2> I don't get this...
<0> salle: heh, no one :)
<1> inviso: I just did... now what do you have to say?
<2> inviso kicked himself by typing that bot command
<2> ?
<0> KimmoA: yeah, forgot the bot doesn't know me yet :)
<12> inviso: Correct. Many people use MyISAM however
<1> salle: help inviso understand that update table set value = value + 125 is actually a read and a write, not some magic silver bullet operation
Return to
#mysql
or
Go to some related
logs:
dd: reading `/dev/hdd': Input/output error ubuntu
ubuntu keeps asking keyring password
#mysql
riceandmeat
#linux
#physics
vga acroynm
administrator mode ubuntu Su returned with an error
irssi list themes
php 5.12
|
|