I have a simple MYSQL question. If I make a query that contains LAST_INSERT_ID() right after an INSERT QUERY running on a web page which has many concurrent users accessing other pages that perform INSERT operations would the value of LAST_INSERT_ID() be adulterated/corrupted?
No, it will return the insert id from the current connection. As long as your script hasn't made any other inserts, you will get the one you want.
Also be aware that this will only return a generated ID (e.g. an auto-increment). If you are creating your own ID's it won't return this to you.
Related
Suppose, that we have 2 databases: a and b, and tables a.test1 and b.test2.
If I need to insert a row into table a.test1, and return LAST_INSERT_ID() to insert into b.test2, will LAST_INSERT_ID() return value from another database? Is it reliable?
I didn't find anything in the manual, but ##IDENTITY depends on client session, so it should be portable between two databases. Isn't it?
LAST_INSERT_ID() always gives you the id of the row inserted by the last INSERT statement you executed on the current connection, irrespective of what table (and what database!) that row went into.
From Mysql documentation:The ID that was generated is maintained in the server on a per-connection basis. This means that the value returned by the function to a given client is the first AUTO_INCREMENT value generated for most recent statement affecting an AUTO_INCREMENT column by that client. This value cannot be affected by other clients, even if they generate AUTO_INCREMENT values of their own. This behavior ensures that each client can retrieve its own ID without concern for the activity of other clients, and without the need for locks or transactions.
Inshort Both *LAST_INSERT_ID()* and *mysql_insert_id()* work as advertised i.e.: they will retrieve the last id inserted into any table during the current session/connection.
I'm using AUTO_INCREMENT and I would like to get that ID of inserted row so that I could update another table using ID as common field between the 2 tables.
I understood that LAST_INSERT_ID will get last ID. However, my concern is that, the database is accessed at same time by many users. Hence, there might be another process accessed the table and also inserted a new row at same time.
Does LAST_INSERT_ID return just the last ID regardless of the connection used, or only return last ID for the connection that I'm using?
Notice, I'm accessing MySQL database using connection pool in Tomcat server.
In summary, I need to insert a row in table A with auto increment, than I need to insert row in table B, which need to be linked to table A using the AUTO_INCREMENT value.
SELECT max(employeeid) FROM Employee;
The above query returns the value of employeeid of last inserted record in Employee table because employeeid is an auto increment column. This seems to be OK, but suppose two threads are executing insert operation simultaneously, there is a chance that you get wrong id of last inserted record!
Don’t worry, MySQL provides a function which returns the value of auto increment column of last inserted record.
SELECT LAST_INSERT_ID();
LAST_INSERT_ID() is always connection specific, this means even if insert operation is carried out simultaneously from different connections, it always returns the value of current connection specific operation.
So you have to first insert record in Employee table, run the above query to get the id value and use this to insert in second table.
LAST_INSERT_ID() work in context, it should be in transactions or inside user defined stored procedures or user defined functions.
LAST_INSERT_ID is connection specific. That's true, but you should be careful if you use connection pooling. This may be problematic when you perform successive INSERT IGNORE statements in a loop and the pool gives you the same connection at each iteration.
For example; Assume that you receive the same (open) connection for each of the below:
INSERT IGNORE ... some-new-id >>> LAST_INSERT_ID returns 100
INSERT IGNORE ... some-existing-id >>> LAST_INSERT_ID still returns 100 (result of the previous operation)
It is always good to check whether the INSERT operation has in fact inserted any rows before calling LAST_INSERT_ID.
LAST_INSERT_ID return the last insert id for the current session.
As long as you don't insert more than one entry with your current connection, it is valid.
Further information here: https://dba.stackexchange.com/questions/21181/is-mysqls-last-insert-id-function-guaranteed-to-be-correct
(i would link to mysql.com, but it'S currently down for me)
I have been pondering on this for a while.
Consider a web application of huge proportions, where, let's say, millions of SQL queries are performed every second.
I run my code:
$q = $db->prepare('INSERT INTO Table
(First,Second,Third,Fourth)
VALUES (?,?,?,?)');
$q->execute(array($first,$second,$third,$fourth));
Then immediately after, I want to fetch the auto incremented ID of this last query:
$id = $db->lastInsertId();
Is it possible for lastInsertId to fail, i.e. fetch the ID of some SQL insert query that was executed between my two code blocks?
Secondary:
If it can fail, what would be the best way to plug this possible leak?
Would it be safer to create another SQL query to fetch the proper ID from the database, just to be sure?
It will always be safe provided that the PDO implementation is not doing something really bone-headed. The following is from the MySQL information on last_insert_id:
The ID that was generated is maintained in the server on a per-connection basis. This means that the value returned by the function to a given client is the first AUTO_INCREMENT value generated for most recent statement affecting an AUTO_INCREMENT column by that client. This value cannot be affected by other clients, even if they generate AUTO_INCREMENT values of their own. This behavior ensures that each client can retrieve its own ID without concern for the activity of other clients, and without the need for locks or transactions.
No. lastInsertId is per-connection, and doesn't require a request to the server - mysql always sends it back in its response packet.
So if the execute method doesn't throw an exception, then you are guaranteed to have the right value in lastInsertId.
It won't ever give you the insert ID of anything else, unless your query failed for some reason (e.g. invalid syntax) in which case it might give you the insert ID from the previous one on the same connection. But not anybody else's.
I am currently working on a big web project using ASP and MySQL.
When inserting into multiple tables I've been using last_update_ID(), but after some research I've found that that SQL statement isn't safe.
So. the problem:
I use two different computers, with different internet connections.
Both computers are logged onto the system I am currently building. I have made a page that prints the connection_id(), and last_update_id.
If I update any table with one of the computers the other one also gets that last_update_ID.
Both computers have the same connection_ID.
What can I do to get around this?
I don't want to (if it's not necessary) do a select statement after the first INSERT; to search for the row that I inserted, to get the correct ID of that row.
It's not my server I am using so I can't make any large changes of the database.
I guess that this problem occurs because the webpages use the same loginName & password to connect to the database, is that true?
Is there any other alternative to get the last update ID? that is totally safe..
I close every connection at the end of the asp page. but that doesn't change the connection_ID.
The connection ID is the for a few minutes even thou I open up different web pages on the server.
I believe the LAST_INSERT_ID() is correct for the current session. So each session receives it's own correct value. Either I don't understand your question or you think you have a problem but you don't.
I am not aware of any LAST_UPDATE_ID() function, on an update you can easily retrieve the updated rows by SELECTing them with the same WHERE clause (before the update)?
reference: http://dev.mysql.com/doc/refman/5.0/en/getting-unique-id.html
For LAST_INSERT_ID(), the most
recently generated ID is maintained in
the server on a per-connection basis.
It is not changed by another client.
It is not even changed if you update
another AUTO_INCREMENT column with a
nonmagic value (that is, a value that
is not NULL and not 0). Using
LAST_INSERT_ID() and AUTO_INCREMENT
columns simultaneously from multiple
clients is perfectly valid. Each
client will receive the last inserted
ID for the last statement that
client executed.
If you want to retrieve the LAST_INSERT_ID from an INSERT query with an ON DUPLICATE KEY UPDATE clause, you can also use the LAST_INSERT_ID() function to retrieve the value of the AUTO_INCREMENT column that was updated:
reference: http://dev.mysql.com/doc/refman/5.0/en/insert-on-duplicate.html
If a table contains an AUTO_INCREMENT
column and INSERT ... UPDATE inserts a
row, the LAST_INSERT_ID() function
returns the AUTO_INCREMENT value. If
the statement updates a row instead,
LAST_INSERT_ID() is not meaningful.
However, you can work around this by
using LAST_INSERT_ID(expr). Suppose
that id is the AUTO_INCREMENT column.
To make LAST_INSERT_ID() meaningful
for updates, insert rows as follows:
INSERT INTO table (a,b,c) VALUES (1,2,3)
ON DUPLICATE KEY UPDATE id=LAST_INSERT_ID(id), c=3;
Your server appears to have connection pooling turned on. What this means is that the database connection is held open after a script finishes, and the next script that comes along uses it, and thus can see any variables that were set on that connection, including LAST_INSERT_ID().
What can't happen is two script instances sharing a connection at the same time. Thus, if your server is busy enough to need to run two script instances at exactly the same time, it will simply create a second database connection, with its own separate LAST_INSERT_ID() variable, and won't interfere with the first.
In short, as long as the INSERT and the LAST_INSER_ID() request happen within the same script (and you don't somehow close the database connection between them), they're completely safe, as your script has exclusive use of that connection.
In mysql, how do I get the primary key used for an insert operation, when it is autoincrementing.
Basically, i want the new autoincremented value to be returned when the statement completes.
Thanks!
Your clarification comment says that you're interested in making sure that LAST_INSERT_ID() doesn't give the wrong result if another concurrent INSERT happens. Rest assured that it is safe to use LAST_INSERT_ID() regardless of other concurrent activity. LAST_INSERT_ID() returns only the most recent ID generated during the current session.
You can try it yourself:
Open two shell windows, run mysql
client in each and connect to
database.
Shell 1: INSERT into a table with an
AUTO_INCREMENT key.
Shell 1: SELECT LAST_INSERT_ID(),
see result.
Shell 2: INSERT into the same table.
Shell 2: SELECT LAST_INSERT_ID(),
see result different from shell 1.
Shell 1: SELECT LAST_INSERT_ID()
again, see a repeat of earlier
result.
If you think about it, this is the only way that makes sense. All databases that support auto-incrementing key mechanisms must act this way. If the result depends on a race condition with other clients possibly INSERTing concurrently, then there would be no dependable way to get the last inserted ID value in your current session.
MySQL's LAST_INSERT_ID()
The MySQL Docs describe the function: LAST_INSERT_ID()
[select max(primary_key_column_name) from table_name]
Ahhh not nessecarily. I am not an MySQL guy but there are specific way to get the last inserted id for the last completed action that are a little more robust than this. What if an insert has happened between you writing to the table and querying it? I know about because it stung me many moons ago (so yeah it does happen).
If all else fails read the manual: http://dev.mysql.com/doc/refman/5.0/en/getting-unique-id.html