Sunday, September 14, 2014

The Windchill Bulk Migrator...


Before I begin, I am wondering that I can see a lot of hits for his topic, but no questions being asked no this topic. Appreciate you inputs/questions here and I will try my level best to  address them.

WBM tool touches almost each and every feature of Windchill. This is one of the latest launched tools by PTC and is used to Migrate from other products to Windchill. It can be also used to migrate from Windchill to Windchill, but honestly speaking, Upgrade Manager works better for Windchill to Windchill upgrade, unless a special case where you might want to upgrade from multiple Windchills to a single Windchill or old Windchill version to an existing Production Windchill.


Recently, I was involved in the WBM project where we migrated to Windchill from another product. The key behind a successful WBM migration project is intensive planning and source data Generation. If you can generate Meta-data of your existing data it comes in very handy. If you can create a meta data which can be seen in the form of a DB table, nothing like it.


So as I was saying, WBM project is planning/research intensive. You need to understand clearly what you have in your source and what do you need to see in Windchill. Sort out all the primary type of objects you have in your source system i.e CAD Docs/drawings, Documents. These objects can be stored on a hard drive. Next are the logical objects like WTParts, CNs etc. These act as primary objects in Windchill but are logical. A non-Windchill system will have primary objects but not necessarily have logical objects with the same mapping as Windchill.


This brings us to the next important step. With what meta-data you have from the source system you should be able to map it to the Windchill objects in the form of Windchill object’s attributes(IBAs).  Once you define a clear attribute map, you can understand which attributes from Windchill are going empty and what needs to defined in them.


Once you have this attribute map, you can go back to Windchill and create all the data that is needed to load the source system data. This can include new Object Subtypes, Users,Products, series etc. This is a non-WBM step and preferred if you can use Windchill Load utility or any other OOTB functionality for all this stuff.

Upon installing WBM software on the Target Windchill, you need to study the entire Staging database to understand what is missing in the Staging DB tables e.g. attributes of objects, extra columns. You need to check if all the attributes for a given object can be loaded using WBM or not, If yes then which table satisfies the need of which table.


Based on these tables you create a table design in which you have to mold the source data so that it can be imported into these tables. Hence it is essential if you can modify your source data into Meta-Data. WBM 4.0 comes with the ease of migrating directly from a Hard Disk. But that’s only for physical objects.

Once you define your physical and logical objects you need to understand and define the links between multiple objects types e.g. EPMBuildRule is used to connect an EPMDoc to a WTPart, Includedin2 is used to connect a Change Notice with its Change Task.


This is the tricky part and hence you need an intensive mapping and more importantly understanding of how you define your source system data.

You cannot just load all the data to WBM and run the loaders all together. The is standard sequence of loading objects to Windchill and totally dependent on what objects you have in hand.

e.g. The first objects always to go in are the EPMDocs and WTDocs since they can survive in Windchill as standalone objects.


I cannot load a Reference link object if there is no corresponding Drawing to it. Hence, we usually will load the physical objects followed by the logical objects followed by the link objects.

WBM is a new tool and you can hit a lot of bumps when working on it and not a lot of data is available on the PTC knowledge base to resolve them. Hence pilot is always recommended so that you can carry out POCs for the different objects for checking data integrity and successful migration with the sourc data in hand.


The sequential progress of a WBM projects looks something like below




Sunday, April 22, 2012

Recoving Oracle..'Wind' database...

It may happen that we might have an Oracle database crash power  down situations and Oracle never comes up.  Oracle might never come up due to corrupt block or any other issues. There may be errors like ORA-01578 or corruption.
There are some simple steps we need to follow
SQL>sqlplus / as sysdba
SQL> shutdown immediate
SQL> startup nomount
SQL>alter database open;
If all these steps go through without any issues then you the lucky person and there are no issues.
In case there is a real corruption we may have to recover the database In this Situation we need to recover the database using REDO logs.
Follow below steps
Take a backup of the entire OCU folder
SQL>sqlplus / as sysdba
SQL> shutdown immediate
SQL> startup nomount

SQL> SELECT MEMBER FROM V$LOG G, V$LOGFILE F WHERE G.GROUP# = F.GROUP#  AND G.STATUS = 'CURRENT';
This query outputs the path of the redo log that we need to submit in the next query.
SQL>recover database until cancel
In this step you should the path the is obtained in the above step
SQL>alter database open resetlogs;
If none of the above work then we need to create a new set of control files in addition to the above steps,
Follow below steps
Take a backup of the entire OCU folder and delete the .CTL files from the OCU folder.
SQL>sqlplus / as sysdba
SQL> shutdown immediate
SQL> startup nomount
SQL>CREATE CONTROLFILE REUSE DATABASE "WIND" RESETLOGS NOARCHIVELOG
    MAXLOGFILES 16
    MAXLOGMEMBERS 3
    MAXDATAFILES 100
    MAXINSTANCES 8
    MAXLOGHISTORY 292
LOGFILE
  GROUP 1 'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDREDO01.LOG'  SIZE 100M,
  GROUP 2 'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDREDO02.LOG'  SIZE 100M,
  GROUP 3 'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDREDO03.LOG'  SIZE 100M
DATAFILE
  'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDBLOBS01.DBF',
  'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDINDEX01.DBF',
  'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDSYSAUX01.DBF',
  'D:\ptc\Windchill_9.1\ocu\oraata\wind\WINDSYSTEM01.DBF',
  'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDUNDOTBS01.DBF',
  'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDUSERS01.DBF',
  'D:\ptc\Windchill_9.1\ocu\oradata\wind\WINDWCAUDIT01.DBF'
;
SQL> SELECT MEMBER FROM V$LOG G, V$LOGFILE F WHERE G.GROUP# = F.GROUP#  AND G.STATUS = 'CURRENT';
This query outputs the path of the redo log that we need to submit in the next query.
SQL>recover database until cancel
In this step you should the path the is obtained in the above step
SQL>alter database open resetlogs;
These all 3 methods usually suffice to recover a crashed database. But if even these 3 methods donot recover your database then you need to get in touch with a database expert.

Tuesday, January 3, 2012

JExplorer..LDAP jobs made easy

The new WindChillDS UI is too complex and boring to work on. I tried to search some tools which could be handy to work on LDAP and found this tool JExplorer.

This tool gives the feel of the old Aphelion browser and has many cool feature that makes it way handy than WindChillDS.

Where to get it…

http://jxplorer.org/downloads/users.html

it is pretty straight forward to use .

Its features;

http://jxplorer.org/features.html

Here is if you need more information,

http://jxplorer.org/documents/JXv3.2_Administrator_Guide.pdf

Tuesday, July 26, 2011

Tell me…what you want to see

I am always thinking of new topic to put up here to help many others who are into Windchill….please reply to this topic below if you want to see any specific topic here that you think might not only help you but several other PLM engineers as well. I will try my best to write an article on it.


Regards,

Rohan

Monday, July 11, 2011

Fiddler to debug Windchill issues on Internet Explorer

Fiddler can be helpful for:

1. Find files/java scripts that are accessed by the client from the server.

2. When we encounter a strange performance work bottleneck in IE that you can't track down.

3. Task is not executable only on Internet Explorer

4. Able to perform a Specific web activity only on Firefox and not on IE

Fiddler works only with IE

What does Fiddler Fiddle with:

Fiddler records all HTTP(S) traffic, set breakpoints, and "fiddle" with incoming or outgoing data.

System Requirements

Windows 2000 / XP / 2003 / Vista with Microsoft .NET Framework v2.0 +
10 megabytes disk space / 800mhz processor (Screams at 2.4ghz)
128 megabytes RAM (512mb+ highly recommended)

Link to Download Fiddler

http://www.fiddler2.com/fiddler2/version.asp

A Picture speaks 1000 words: Fiddler Layout




HTTP Statistics view

By exposing all HTTP traffic, Fiddler easily shows which files are used to generate a given page. Using the Statistics page, the user can multiselect to get a "total page weight"—the number of requests and the bytes transferred


Additionally, by exposing HTTP Headers in the Session list, the user can see whether pages are missing HTTP Expiration headers that permit client or proxy caching. If a response does not contain Expires or Cache-Control headers, it might not be cached by the client.

Using Fiddler for Debugging

In addition to seeing all HTTP requests and responses, Fiddler supports the notion of breakpoints. When the Enable Single Step Debugging option is checked on the Rules menu, or when the properties of the HTTP Request or Response match the target criteria, Fiddler can pause HTTP traffic and allow edits. This feature proves useful for security testing, as well as for general functionality testing, because all code paths can be exercised.