Skip to main content
3-Newcomer
November 7, 2025
Question

Windchill application restart is particularly slow

  • November 7, 2025
  • 3 replies
  • 302 views
I am using Windchill PDMLink Release 13.0 and Datecode with CPS 13.0.1.2

Windchill application restart is particularly slow, taking about 500 seconds. Which aspects should be optimized to improve this?

    3 replies

    avillanueva
    23-Emerald I
    23-Emerald I
    November 7, 2025

    Check that the database is in working order. Profiling it during startup can help spot any long running queries that could be being executed. Can you tell if its timing out or stuck for a period at a certain step? You should ensure that the system is tuned properly based on the Windchill configurator (WCA). It should have the right amount of memory to start up and operate. 

    16-Pearl
    November 7, 2025

    Additionally....
    Anti-virus can add a lot of overhead/delay to starting Windchill if it wasn't configured for Windchill.

    Infrastructure issues that show as poor disk IO can really slow down startup.

    Tomcat scans codebase for jar files on startup.  Poor disk performance can make this a VERY long process.

    https://www.ptc.com/en/support/article/CS344564

    avillanueva
    23-Emerald I
    23-Emerald I
    February 24, 2026

    @HX_11259932, were you able to reduce start up time? Let us know what you did.

    18-Opal
    February 25, 2026

    What's interesting is - I saw this issue when I was doing initial testing for upgrading to WC 13.0 from WC 12.0 before I joined PTC. It seemed to be an issue until I went to testing the production environment, where the issue went away. We had a 3 tier environment (DEV, Staging, Prod). DEV and Staging both saw symptoms of this where startup time was much longer than the previous comparable environment (Dev 12 to Dev 13, STG 12 to STG 13). Staging was set up to mirror what our production system would look like including resources. And it still started slowly. On the production VM, we didn't see the issue... and strangely enough, by the time we got to testing upgrades with what would be the prod WC 13 machines, the issues were fixed.

     

    One thing to maybe look at is what does your security software do? We used a heuristic scanning tool (Sentinel One) and by the time we got to production, we had exceptions built in to address issues found in testing. Tools like Sentinel One inject themselves in EVERY java process unless excepted. We had many anomalies with this, that caused all sorts of performance issues