From 7c749e17f1711d7cddc545bc0d49fe915b7417de Mon Sep 17 00:00:00 2001 From: eunmoliveira Date: Tue, 6 Jan 2026 15:21:06 +0000 Subject: [PATCH 1/2] exit-codes-disclaimer added on FDS and Delft3D FAQ --- docs/07.delft3d/2.faq/0.faq.md | 20 +++++++++++++++++++- docs/10.fds/4.faq/0.faq.md | 22 ++++++++++++++++++++-- 2 files changed, 39 insertions(+), 3 deletions(-) diff --git a/docs/07.delft3d/2.faq/0.faq.md b/docs/07.delft3d/2.faq/0.faq.md index f4aa4c3..580101c 100644 --- a/docs/07.delft3d/2.faq/0.faq.md +++ b/docs/07.delft3d/2.faq/0.faq.md @@ -6,7 +6,25 @@ seo: description: Find answers to commonly asked questions about Delft3D. --- -## 1. How do I run a Delft3D simulation with FLOW-WAVE coupling? +## 1. Why is my Delft3D FM simulation marked as “Success” even when it fails? +Inductiva determines task status based on the program’s exit code. Typically, when Unix programs finish, they return an exit code indicating how the program terminated: + +- **Exit code 0** → the program completed successfully (Inductiva marks the task as “Success”) +- **Non-zero exit code** → an error occurred (Inductiva marks the task as “Failed”) + +Most simulators follow this convention. However, in some error cases, **Delft3D FM returns an exit code of 0 even though the simulation has failed**. As a result, Inductiva may incorrectly mark the task as "Successful". + +This does **not** mean that Delft3D FM is malfunctioning on Inductiva. The issue is limited to how certain failure conditions are reported. + +### What is being done about this? +We’re evaluating several approaches to address this behaviour. + +### What should I do for now? +Please check the simulation logs (`stdout.txt` and `stderr.txt`) to confirm whether a run succeeded or failed, rather than relying solely on the task status. + +
+ +## 2. How do I run a Delft3D simulation with FLOW-WAVE coupling? To run a Delft3D simulation with FLOW-WAVE coupling using the Inductiva API, you’ll need to run both the FLOW and WAVE components together. This involves creating a shell script (e.g., `run_sim.sh`) that starts `d_hydro.exe` with diff --git a/docs/10.fds/4.faq/0.faq.md b/docs/10.fds/4.faq/0.faq.md index 9e85d9e..6b0b124 100644 --- a/docs/10.fds/4.faq/0.faq.md +++ b/docs/10.fds/4.faq/0.faq.md @@ -6,7 +6,25 @@ seo: description: Find answers to commonly asked questions about FDS. --- -## 1. Why aren’t live logs showing up in the Web Console? +## 1. Why is my FDS simulation marked as “Success” even when it fails? +Inductiva determines task status based on the program’s exit code. Typically, when Unix programs finish, they return an exit code indicating how the program terminated: + +- **Exit code 0** → the program completed successfully (Inductiva marks the task as “Success”) +- **Non-zero exit code** → an error occurred (Inductiva marks the task as “Failed”) + +Most simulators follow this convention. However, in some error cases, **FDS returns an exit code of 0 even though the simulation has failed**. As a result, Inductiva may incorrectly mark the task as "Successful". + +This does **not** mean that FDS is malfunctioning on Inductiva. The issue is limited to how certain failure conditions are reported. + +### What is being done about this? +We’re evaluating several approaches to address this behaviour. + +### What should I do for now? +Please check the simulation logs (`stdout.txt` and `stderr.txt`) to confirm whether a run succeeded or failed, rather than relying solely on the task status. + +
+ +## 2. Why aren’t live logs showing up in the Web Console? FDS behaves differently from most simulators: it writes all logs to `stderr` instead of `stdout`. Since the Web Console only streams `stdout`, no logs will appear there during the simulation. @@ -21,7 +39,7 @@ for viewing in the web console.
-## 2. Why can’t I use more `n_vcpus` for my simulation? +## 3. Why can’t I use more `n_vcpus` for my simulation? FDS parallelizes simulations by assigning each mesh to a separate vCPU. If your simulation defines only a single mesh, it can only use one vCPU. Attempting to use more will result in an error. For example, the following setup allows only **1 vCPU**: From ad001cc7adecbf8edfcfd0edd99163fd5b307dbe Mon Sep 17 00:00:00 2001 From: eunmoliveira Date: Tue, 6 Jan 2026 18:44:35 +0000 Subject: [PATCH 2/2] update-faqs-final --- docs/07.delft3d/2.faq/0.faq.md | 6 +----- docs/10.fds/4.faq/0.faq.md | 6 +----- 2 files changed, 2 insertions(+), 10 deletions(-) diff --git a/docs/07.delft3d/2.faq/0.faq.md b/docs/07.delft3d/2.faq/0.faq.md index 580101c..e4b2796 100644 --- a/docs/07.delft3d/2.faq/0.faq.md +++ b/docs/07.delft3d/2.faq/0.faq.md @@ -12,14 +12,10 @@ Inductiva determines task status based on the program’s exit code. Typically, - **Exit code 0** → the program completed successfully (Inductiva marks the task as “Success”) - **Non-zero exit code** → an error occurred (Inductiva marks the task as “Failed”) -Most simulators follow this convention. However, in some error cases, **Delft3D FM returns an exit code of 0 even though the simulation has failed**. As a result, Inductiva may incorrectly mark the task as "Successful". +Most simulators follow this convention. However, in some error cases, **Delft3D FM returns an exit code of 0 even though the simulation has failed**. As a result, Inductiva may misclassify the task as "Successful". This does **not** mean that Delft3D FM is malfunctioning on Inductiva. The issue is limited to how certain failure conditions are reported. -### What is being done about this? -We’re evaluating several approaches to address this behaviour. - -### What should I do for now? Please check the simulation logs (`stdout.txt` and `stderr.txt`) to confirm whether a run succeeded or failed, rather than relying solely on the task status.
diff --git a/docs/10.fds/4.faq/0.faq.md b/docs/10.fds/4.faq/0.faq.md index 6b0b124..c7849ae 100644 --- a/docs/10.fds/4.faq/0.faq.md +++ b/docs/10.fds/4.faq/0.faq.md @@ -12,14 +12,10 @@ Inductiva determines task status based on the program’s exit code. Typically, - **Exit code 0** → the program completed successfully (Inductiva marks the task as “Success”) - **Non-zero exit code** → an error occurred (Inductiva marks the task as “Failed”) -Most simulators follow this convention. However, in some error cases, **FDS returns an exit code of 0 even though the simulation has failed**. As a result, Inductiva may incorrectly mark the task as "Successful". +Most simulators follow this convention. However, in some error cases, **FDS returns an exit code of 0 even though the simulation has failed**. As a result, Inductiva may misclassify the task as "Successful". This does **not** mean that FDS is malfunctioning on Inductiva. The issue is limited to how certain failure conditions are reported. -### What is being done about this? -We’re evaluating several approaches to address this behaviour. - -### What should I do for now? Please check the simulation logs (`stdout.txt` and `stderr.txt`) to confirm whether a run succeeded or failed, rather than relying solely on the task status.