We have a Companies table in our database with a "Name" varchar column and its size is currently 30 characters. Our seemingly simple task of changing its size to 50 characters has turned into a bit of an issue. When trying to change it and save the changes through Sybase Central we get the following error:
[Sybase][ODBC Driver][SQL Anywhere]Illegal column definition: Name
SQLCODE: -1046
SQLSTATE: 42000
SQL Statement: ALTER TABLE "DBA"."Companies" ALTER "Name" VARCHAR(50)
We've tried various escaping characters around the column thinking it might have something to do with the word "Name" being treated differently internally by Sybase. We have no indexes or constraints on this column and have removed the single trigger that did exist just trying to isolate any potential factors. Further perplexing us, we have a Companies_a table that keeps track of all changes to the Companies table that has nearly the exact same schema including the Name column. We are able to change that column's size without issue which seems to indicate it's not necessarily an issue with the word "Name". I've gone through all the tabs in Sybase Central for this table, and don't see anything special/different about this table or this column.
Googling this issue is difficult as the word "Name" is extremely common. We have workarounds we can do (ie. creating temp column, copying, dropping & recreating the column, copying back) but if possible I'd like to understand exactly what's happening here and why.
I believe our Companies table and this column were likely created in a previous version of ASA. The inline_max column in sys.systabcol was set to null and we could not change it.
Running the following statement set the defaults:
ALTER TABLE "DBA"."Companies" ALTER "Name" inline use default prefix use default;
I was then able to run this statement without error:
ALTER TABLE "DBA"."Companies" ALTER "Name" VARCHAR(50);
Here's where I found the general concept for the fix:
Sybase SqlAnywhere forum thread
ALTER TABLE [tableName]
ALTER [ColumnName] varchar(4000) NULL
Related
I've come from MS SQL Server where there's an option in SSMS to allow the dropping of tables/views if required when altering them.
Is there the same option in pgAdmin for PostgreSQL?
When trying to change the type of an existing column in a table, I keep getting the error:
ERROR: cannot alter type of column used by a view or rule .... on view ... depends on column "name"
FWIW I'm changing the type from text to character varying, so it's hardly magician work
This one is most odd, I've got a DB2 instance with 50+ tables defined, and whilst I can insert and query data. DB2 is being extremely picky about formatting and keeps complaining about both table / column context whilst insisting on everything being quoted.
Most weird is that none of the tables show in the results of a 'list tables' command whilst 2 other tables defined by API do..?
Syntax I used to create the tables..
CREATE TABLE Shell.Customers
(
"idCustomers" BIGINT NOT NULL GENERATED ALWAYS AS IDENTITY ( INCREMENT BY 1 NO CYCLE ORDER ),
"Name" VARCHAR(64) NOT NULL,
"Code" VARCHAR(6) NOT NULL,
PRIMARY KEY ("idCustomers")
) COMPRESS YES ADAPTIVE WITH RESTRICT ON DROP;
Any ideas where I messed it up?
Thanks in advance.. :)
LIST TABLES command without ‘FOR’ clause shows tables for the current user. Your table is not listed unless your current user name is SHELL.
Use LIST TABLES FOR SCHEMA SHELL (or FOR ALL) command to list the table you mentioned.
I'm trying to add a unique constraint on 2 columns on DB2-- docType column and title column. The values of each column can repeat in themselves. however, the values of (docType, title) pairs should not repeat.
I tried so far
ALTER TABLE externalfiles
ADD CONSTRAINT logicalKey UNIQUE (doctype, title)
and
alter table externalfiles add unique (doctype, title)
, and got the following error to both:
Operation not allowed for reason code "7" on table "PIT.EXTERNALFILES".. SQLCODE=-668, SQLSTATE=57016, DRIVER=4.21.29
Isn't this any allowed??
DB2 Unique Constraint over multiple Columns suggests creating indexes as alternative to this. Haven't tried indexing them yet-- however, i'm wondering why the unique constraint isn't working.
TIA.
This web search on db2 luw sqlstate 57016 returns a number of links that suggest a prior ALTER may require that a REORG be performed before the ADD CONSTRAINT can be effected.
Perhaps most notable was the following doc link and snippet of text describing the RC7 [apparently SQLERRMC=7]:
DB2 for Linux UNIX and Windows 9.7.0
->Database fundamentals
->Messages
->SQL Messages
->SQL0500 - SQL0999
->SQL0668N
…
7 • The table is in the reorg pending state. This can occur after an
ALTER TABLE statement containing a REORG-recommended operation.
…
First try to reorganize table by
call sysproc.admin_cmd ('reorg table <TABLE_NAME>');
Than try again to create unique index. It should work.
I was simply wondering if there was a performance or other technical reason for you to be unable to perform a ALTER TABLE ALTER COLUMN statement with multiple columns within the same line e.g.
ALTER TABLE tblGeneric ALTER COLUMN Generic1 VARCHAR(255), Generic2 VARCHAR(255);
This is exclusively a restriction of t-sql as you can in fact comma separate columns with the MODIFY statement of mysql.
I just thought it odd, especially considering the MODIFY of mysql, that you can do a same line multi ALTER TABLE ADD statement but not a same line multi ALTER. I was just wondered if there is any particular documented reason for this or at least if it's in a issues list.
SQL Server T-SQL doesn't allow multiple columns to be changed in one ALTER TABLE command (unlike some other languages where it's possible).
Please follow this MSDN link for ALTER command syntax and explanation.
However, you can do multiple ADD or multiple DROP COLUMN, but just one ALTER COLUMN.
I need to increase the size of a character varying(60) field in a postgres database table without data loss.
I have this command
alter table client_details alter column name set character varying(200);
will this command increase the the field size from 60 to 200 without data loss?
The correct query to change the data type limit of the particular column:
ALTER TABLE client_details ALTER COLUMN name TYPE character varying(200);
Referring to this documentation, there would be no data loss, alter column only casts old data to new data so a cast between character data should be fine. But I don't think your syntax is correct, see the documentation I mentioned earlier. I think you should be using this syntax :
ALTER [ COLUMN ] column TYPE type [
USING expression ]
And as a note, wouldn't it be easier to just create a table, populate it and test :)
Yes. But it will rewrite this table and lock it exclusively for duration of rewriting — any query trying to access this table will wait until rewrite finishes.
Consider changing type to text and using check constraint for limiting size — changing constraint would not rewrite or lock a table.
you can use this below sql command
ALTER TABLE client_details
ALTER COLUMN name TYPE varchar(200)
From PostgreSQL 9.2 Relase Notes E.15.3.4.2
Increasing the length limit for a varchar or varbit column, or removing the limit altogether, no longer requires a table rewrite.
Changing the Column Size in Postgresql 9.1 version
During the Column chainging the varchar size to higher values, table re write is required during this lock will be held on table and user table not able access
till table re-write is done.
Table Name :- userdata
Column Name:- acc_no
ALTER TABLE userdata ALTER COLUMN acc_no TYPE varchar(250);