Skip to main content
14-Alexandrite
July 27, 2016
Solved

Windchill log tool

  • July 27, 2016
  • 3 replies
  • 6441 views

Does anyone know of a good tool to view Windchill logs? I heard Apache Chainsaw would work, but do not see any documentation to support this.

Best answer by jessh

You do know about http://yourHostAndPort/Windchill/wtcore/jsp/jmx/logFiles.jsp, right?

That provides an HTML UI for searching/browsing all the log files from the cluster.

Also, you know about http://yourHostAndPort/Windchill/wtcore/jsp/jmx/logEvents.jsp, right?

It provides an ability to search all "persisted" log events via the database.  These are essentially the warnings and errors (captured by a custom log4j appender, AsyncJDBCAppender).  The difference is that

  1. This uses database queries
  2. The log event data here is more structured and complete than that in the log file in cases

This is all assuming you're on 10.0 or later, of course.  These capabilities (and all those accessible from https://yourHostAndPort/Windchill/wtcore/jsp/jmx/index.jsp) continue to improve in each release and MOR up through 11.0.

Of course, these are not the end-all-and-be-all, and, yes, you can configure socket appenders and set up chainsaw if you like -- or use Splunk or various other tools.  It's just log4j and log files (apart from the custom database appender).

3 replies

jessh12-AmethystAnswer
12-Amethyst
July 28, 2016

You do know about http://yourHostAndPort/Windchill/wtcore/jsp/jmx/logFiles.jsp, right?

That provides an HTML UI for searching/browsing all the log files from the cluster.

Also, you know about http://yourHostAndPort/Windchill/wtcore/jsp/jmx/logEvents.jsp, right?

It provides an ability to search all "persisted" log events via the database.  These are essentially the warnings and errors (captured by a custom log4j appender, AsyncJDBCAppender).  The difference is that

  1. This uses database queries
  2. The log event data here is more structured and complete than that in the log file in cases

This is all assuming you're on 10.0 or later, of course.  These capabilities (and all those accessible from https://yourHostAndPort/Windchill/wtcore/jsp/jmx/index.jsp) continue to improve in each release and MOR up through 11.0.

Of course, these are not the end-all-and-be-all, and, yes, you can configure socket appenders and set up chainsaw if you like -- or use Splunk or various other tools.  It's just log4j and log files (apart from the custom database appender).

12-Amethyst
July 28, 2016

P.S. Out-of-the-box most warnings are not captured to log files, whereas they are persisted to the database, so that's one big difference between the log files and the persisted log event data.

tgudobba14-AlexandriteAuthor
14-Alexandrite
July 28, 2016

Thanks Jess, we were looking for a way to read the logs easier, and this looks like it will do the trick!

12-Amethyst
July 29, 2016

P.S. If the Monitoring Tools pages don't help you mine the persisted health/performance/logging data well enough, you can always either:

  1. Configure to enable freeFormQuery.jsp (in recent releases).  There should be docs on this (else look in the JSP and also see create_wc_perf_user.sql in db/sql).
  2. Write your own JSP (or whatever) to query the database tables in question.  Do note that out-of-the-box large strings in these tables are stored in BLOBs, not CLOBs to allow compression.  While you could alter the schema for the non-modeled tables in question to use CLOBs (the code should all continue to work either way), there are utility methods one can use to handle the BLOBs -- see the existing JSPs for details.
15-Moonstone
September 14, 2016

Tim Gudobba‌‌ & Jess Holle‌,

Just to throw in my 2 cents...  I find it helpful and beneficial "tail" the log files in the terminal Window - especially when I'm doing development work on a test/dev server.  I'm not sure if you're running Windows or Unix, but we're running Unix.  The "tail" command can do that.  That way, as new information is written to the logs, it is also shown live in the terminal window.

We actually took it step further and put a custom script around the tail command.  The script tails both of our MS and our BMS simultaneously.  That way we always see the message, no matter which MS the message is written to.