In article <[EMAIL PROTECTED]>,
"Aleksandr V. Dyomin" <[EMAIL PROTECTED]> writes:
> $key='somekeyvalue';
> dbquery("update sometable set count=count+1 where keyfield='$key'");
> if(mysql_affected_rows()<1)
> dbquery('insert into sometable set keyfield='$key', count=1');
> ---
> First question: this is good method?
It's good if you expect the UPDATE normally to succeed. Otherwise,
you should first try the INSERT.
> Second... My script work on many different hosts with different
> hardware, os(only Linux or FreeBSD), and different PHP and MySQL
> version. It works fine excepting one thing... Sometime happens errors
> like:
> MySQL: Duplicate entry 'somekeyvalue' for key 1, query was: insert into
> sometable set keyfield='somekeyvalue', count=1
> Why this happens? On some hosts this messageis very rare, but on others
> - so often... I cant understand reason :(
You have a race condition:
* Client 1 tries UPDATE, sees that it fails
* Client 2 tries UPDATE, sees that it fails
* Client 1 does INSERT - okay
* Client 2 does INSERT - duplictae key error
If you think this happens seldom, do the following:
1. Try UPDATE
2. If it fails: try INSERT
3. If it fails due to a duplicate key error: repeat step 1
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
