You are not logged in.
- Topics: Active | Unanswered
Pages: 1
#1 Re: mORMot 2 » Binding multiple ports to multiple routes » 2026-07-13 14:25:39
So you want a single THttpApiServer, but several ports?
Just one question: why several ports?
No, to be exact, different ports for different paths in THttpAsyncServer.
My use case: It is served as an HTTP server that receives requests from various SDK servers. Some SDKs require that the URL must be the root path, which is beyond my control. Currently, I use THTTPAPIServer, which works quite well for this scenario. However, I want to migrate to the new asynchronous HTTP server.
Each SDK's individual network traffic is not very large, but when added together, it becomes substantial. I do not want to create multiple THTTPServer instances for every single SDK, as that might create many threads that I do not need.
It is a common feature that supported by Nginx or IIS.
#2 Re: mORMot 2 » Binding multiple ports to multiple routes » 2026-07-06 15:06:56
If I need to use an async handle to reponse the request, it seems my only option is to setup an Nginx server as a reverse proxy, as I cannot setup multiple HTTP server to handle the logic as the non-async version could do.
@ab: Is it possible this could be supported?
#3 Re: mORMot 2 » Compiling mORMot 2 for Linux with Delphi's LLVM compiler: Is it feasib » 2026-03-27 16:41:36
Supporting Delphi for Linux support is fantastic! 
#4 Re: mORMot 2 » Binding multiple ports to multiple routes » 2026-03-27 16:32:12
Can you explain a little bit more?
Sure.
When I use THttpApiServer, I can add multiple routes with different ports using AddUrl.
For example:
AddUrl('test1','8056')
AddUrl('test2','8097')
But when I am using THTTPServer, it seems I can only pass the port parameter to the constructor, it does not support to bind different routes to different ports.
#5 mORMot 2 » Binding multiple ports to multiple routes » 2026-03-26 06:35:28
- mormot.coremy_only_lonely
- Replies: 5
I'm using THttpApiServer to achieve this goal in the past, which works great.
But now I'm tring to port the application to Linux, so THttpApiServer is no longer an option.
Do I have to create multiple THttpServer instance to achieve the same goal?
#6 Re: mORMot 2 » Compiling mORMot 2 for Linux with Delphi's LLVM compiler: Is it feasib » 2026-03-21 14:08:35
It seems the first step is done:
https://github.com/synopse/mORMot2/comm … 2d45a29e61
#7 Re: mORMot 2 » Compiling mORMot 2 for Linux with Delphi's LLVM compiler: Is it feasib » 2026-03-03 14:06:20
I think the whole community will welcome this change!
#8 Re: mORMot 2 » Compiling mORMot 2 for Linux with Delphi's LLVM compiler: Is it feasib » 2026-02-26 14:13:55
So did I. Most of the code just work after changing platforms, except for the parts relying on Windows unit. For us, performance is already good enough; compatibility is our main concern.
#9 Re: mORMot 2 » Compiling mORMot 2 for Linux with Delphi's LLVM compiler: Is it feasib » 2026-02-25 16:59:11
Will Delphi for Linux support still be considered in the roadmap?
Surely we could use FPC for new projects. But a lot of people like me are trying to migrate the old Windows project to Linux, a total rewrite in FPC is often not practical. To be honst, I have used a lot of 3rd party libraries in my project, but for now mORMot is the last 3rd party library that does not support Delphi for Linux. I have to write a adapter to use mORMot and Indy version of HTTP Server.
#10 mORMot 2 » Any plans to support WebTransport Server? » 2025-03-04 16:19:37
- mormot.coremy_only_lonely
- Replies: 1
Any plans to support WebTransport Server?
https://developer.mozilla.org/en-US/doc … bTransport
#11 Re: mORMot 2 » Continue discusstion Github issue #321 (Race condition in THttpAsyncS) » 2024-12-18 14:04:29
Do we have any progress on this? I see some related commits, but the issue we discussed is not fixed.
#12 Re: mORMot 2 » Continue discusstion Github issue #321 (Race condition in THttpAsyncS) » 2024-12-11 16:02:44
But switching cores might not be an instant behavior.
#13 Re: mORMot 2 » Continue discusstion Github issue #321 (Race condition in THttpAsyncS) » 2024-12-11 02:47:30
That's great!
My thought is always, correctness first, performance later.
This could also happen when only one or some cores are 100% usage, not all cores.
#14 mORMot 2 » Continue discusstion Github issue #321 (Race condition in THttpAsyncS) » 2024-12-09 15:33:05
- mormot.coremy_only_lonely
- Replies: 6
The Github issue was closed. But I still think there are some race conditions.
Pages: 1