TotalCAE now offers an industry first capability of examining running job output for a wide variety of application specific problems we have seen in our twenty years of running CAE applications, and proactively alerts the engineer to investigate through email and our portal.
With on-premise clusters, the user historically would just notice at some point that the model was not working as they desired, fix it and resubmit. However with the advent of public cloud, if a model suddenly is stuck, not progressing/converging, you are paying for cloud compute time that may not being put to useful work.
With CAEWatchDog, you can be immediately notified something is looking amiss with the model, a bug in the solver, or other application specific issues so the engineer can decide to cancel the job, or take some remediation to save time, and real money when running on the cloud.
CAEWatchdog has application specific detection, for example in LS-DYNA it can detect LS-DYNA specific issues that can occur like
1. Error (or Normal) termination in the model, but LS-DYNA did not exit.
2. L2A taking too long due to model issues.
3. No new d3plots written in an abnormally long time indicating model issue.
4. License hanging issues.
and more.
TotalCAE CAEWatchdog is included in both our SMB and Enterprise management plans both on-premise and on TotalCAE Public cloud.