Fixing a corrupt project- Step by Step

[hamzaqk]

I know there are alot of post here pertaining to project corruption but most of them just point to another post and when you go on there, there is no solid solution to the problem. So would someone or all of us be kind enough to constibute to this one and post thier experiences on this so the information is avaliable at one place: i have tried the following so far:

Error:

Code: Select all

Error calling subroutine: DSR_PROJECT (Action=1); check DataStage is set up correctly in project dwhdrop3stgdev 
(Unable to load subroutine on host (30105)) 

Step 1

Code: Select all

 LOGOT PROJECT NAME 

Step2
First thing i checked for the schema by doing this

Code: Select all

VERIFY.SQL SCHEMA PROJECTNAME

the above gave me error

Code: Select all

 ** No SQL catalog data  for schema ' 'in UV_SCHEMA

Step 3
then i run the commans to rebuild the index by:

Code: Select all

 DS.REINDEX ALL 

Step 4 i run the fix command

Code: Select all

 VERIFY.SQL SCHEMA FIX 

Unfortunately this does'nt solve the problem either.

I dont know what step 5 should be now?

[ray.wurlod]

Begin by determining that there is an entry for the project in both the UV.ACCOUNT and UV_SCHEMA system tables.

Code: Select all

SELECT @ID,PATH FROM UV.ACCOUNT WHERE @ID = '<<ProjectName>>';

Code: Select all

SELECT SCHEMA_NAME, PATH FROM UV_SCHEMA WHERE SCHEMA_NAME = '<<ProjectName>>';

Let's make sure that your Step 1 is correct next.

Code: Select all

LOGTO ProjectName

Was this successful? Was there an error message? (You mis-spelled the command.)
Did you actually end up in the project? The command WHO executed at the TCL prompt will report which account you are in. You should secondarily verify that you are in the correct directory.

Code: Select all

DOS /C CD

(UNIX users should use SH -c pwd here.)

If the command WHO reports UV rather than the project name, then the LOGTO was not successful, which might well be the case if the project is corrupted.

In that case you need to determine the contents of the project directory. At the operating system level obtain a directory listing of the project directory's contents. Please post that listing here (in Code tags); it will not show any confidential information.

Based on the contents of the project directory, and of entries in the UV.ACCOUNT and UV_SCHEMA tables, you can then determine whether VERIFY.SQL will be of any use to you. As you've seen, arbtrarily requesting a FIX is not sufficient - there are various syntax options needed to tell the utility more precisely what to do.


To answer your main question, there is no single solution to a corrupted project. It depends how and why it got corrupted.

[ray.wurlod]

Let's assume that the project directory exists, and that there is a correct entry in UV.ACCOUNT for it but no entry for it in UV_SCHEMA. To determine whether that is the case, use VERIFY.SQL with the SCHEMA keyword, the pathname of the project directory (note: NOT the project name) and without the FIX option.

Code: Select all

VERIFY.SQL SCHEMA E:/Ascential/DataStage/Projects/MyProject

(I have gotten into the habit of using "/" rather than "\" in Windows pathnames with DataStage utilities - some of the DataStage utilities seem to prefer it and, in most cases, Windows is happy with either.)