Forum Replies Created
-
AuthorPosts
-
peterlaursen
Participantok .. maybe I misunderstood.
I can use a alias.dyndns.dk adress (and I use it to connect to servers running on virtual machines – the only solution I found here as also I normally have a dynamic ip adress) for the machine where the server is running, but I never tried creating a user with a dyndns host specifier. I am not sure if it works.
Did you enter the FULL subdomain as host specifier when defining user like 'john'@'john.dyndns.org' ?
peterlaursen
ParticipantIf your client machine is having dynamic ip you can use wildcards like '197.211.3%.%'. You may ask your hosting provider for the range of their ip-pool – or find out by experience.
I also think you will be able to use a dynamic DNS service (like dyndns.dk).
peterlaursen
ParticipantHow big a chunk size have you selected?
The chunk size defines how much is transferred from the server, the MB setting has effect only after the log chunck was transferred.
Can you monitor your network traffic and see what is happing here? Are data being transferrred?
peterlaursen
ParticipantBTW: If you are coding for Windows/.NET then MySQL has a plugin for Visual Studio.
I never tried it myself, but saw problems reported with other VS versions than 2005.
Documentation: http://dev.mysql.com/doc/refman/5.0/en/con…or-vstudio.html
Go to the bottom of this FTP-page for download: ftp://mirrors.dotsrc.org/mysql/Downloads/Connector-Net/
peterlaursen
ParticipantNope! I have to correct myself (I did a mistake). For me both 'copy to other' and 'duplicate table' do not touch cardinality with the few examples I tried till now. But we will try with different table ENGINES, different server versions etc.
Can you share the table? If so please zip and attach!
And please tell SQLyog version and server versions!
peterlaursen
ParticipantI can confirm there is an issue .. also with 'duplicate table'. And for me ANALYZE TABLE does not do the job with server 5.0.62. We need to research into this!
peterlaursen
Participantanyway .. please tell how you copied.
if SQLyog 'copy to other' leaves cardinality wrongly NULL we should possibly invoke ANALYSE TABLE after copy!
peterlaursen
ParticipantPlease read too:
http://dev.mysql.com/doc/refman/5.0/en/show-index.html
Quote:CardinalityAn estimate of the number of unique values in the index. This is updated by running ANALYZE TABLE or myisamchk -a. Cardinality is counted based on statistics stored as integers, so the value is not necessarily exact even for small tables. The higher the cardinality, the greater the chance that MySQL uses the index when doing joins.
So if some SHOW statement reports cardinality different from what you expect first try ANALYZE TABLE. But be aware that in MySQL cardinality is *estimated* only!
peterlaursen
ParticipantMySQL is a database manager for MySQL.
We have no plans for making SQLyog a tool for *direct application programming*.
peterlaursen
ParticipantHOW did you 'transfer table to another server'? With SQLyog 'copy to other' or some other way uisng some other tool?
Did you transfer both structure and data?
What ENGINE does the table use (MyISAM, InnoD:cool:?
Were indexes created on 'the other server'?
Is cardinality NULL on the 'first' or the 'otehr' server?
Plese describe the problem *a reproducable way* !
The way you ask nobody can understand the problem!
peterlaursen
ParticipantThe most simple reply to the question would be SQL like (using `test`database)
Code:SELECT
subcategory.id
, subcategory.category
, masterlist.sub_cat
— and so on for more columns from masterlist
FROM
test.subcategory
JOIN test.masterlist
ON (subcategory.id = masterlist.sc_id);but you may want a WHERE, ORDER BY etc. too of course
Refer to
http://dev.mysql.com/doc/refman/5.0/en/join.html
for full JOIN syntax
SQLyog Enterprise has a GUI Query Builder. Here the above example will look like attached image (more usage details described in help file).
[attachment=964:qb.jpg]
peterlaursen
ParticipantAlso an tracker report to support Adarsh's request was added:
peterlaursen
ParticipantAdded to our issue-tracker here:
peterlaursen
ParticipantThis is quite another discussion than the one raised by Adarsh. This will not require any parsing of the content of the SQL file. We would only need to invoke the refresh routine for the active node in the Object Browser after import. And I think this will not be hard to do! And obviously you are right that if for instance 'tables' a database is expanded in the OB, then new tables may be added with the import (for instance).
peterlaursen
Participant1) all date/time datatypes in MySQL are stored as strings. No FLOAT problem here!
2) we do not consider UNIQUE KEYS when updating from the GRIDS. One reason is that a UNIQUE KEY may be defined on NULLABLE columns and for those columns where the value of such column IS NULL a UNIQUE KEY is actually non-unique!
3) I think the native date/time datatypes will be fastest and also most convenient in most situations. Only for data where it could happen that a DATE is TOO OLD for the MySQL DATE etc. type (like the year where Zarathustra was born!) types I would consider storing year, month, day etc. as seperate INTEGER values. However storing a UNIX_TIMESTAMP (as integer or as string) is also always a possibility. This is widely used by standard web applications designed to work on different RDBMS as UNIX_TIMESTAMP type is the same everywhere as opposite to DATE, TIME, DATETIME, TIMESTAMP etc. being implemented differently in different RDBMS. An example is our own FAQ system (phpMyFaq) that runs on many different RDBMS (MySQL, MS SQL, Oracle, Sybase, DB-2, SQLite and probably more)
Functions to work with UNIX_TIMESTAMPS are described here: http://dev.mysql.com/doc/refman/5.0/en/dat…-functions.html
4) I am not able to understand why updation failed for you as there are no FLOAT types. Could you provide an example (just a small table with a few rows or at least a structure-only DUMP) where this is reproducable? Maybe some detail that we should fix/consider/document. Also: did you check in HISTORY what SQL was sent?
-
AuthorPosts