SyliusResourceBundle
SyliusResourceBundle copied to clipboard
Sylius
Asynchronous streaming communication for Clojure - web server, web client, and raw TCP/UDP
When rejecting an invalid request, we close the underlying connection [here](https://github.com/ztellman/aleph/blob/be1ee04c130660ce69ee7c6603132b36be79ec1f/src/aleph/http/server.clj#L211-L219). But when rejecting it because of `RejectedExecutionException`, we [do not](https://github.com/ztellman/aleph/blob/be1ee04c130660ce69ee7c6603132b36be79ec1f/src/aleph/http/server.clj#L160-L168). We probably do not need to do this, when...
HTTP pipelining support is [not obvious](https://github.com/ztellman/aleph/blob/be1ee04c130660ce69ee7c6603132b36be79ec1f/src/aleph/http/server.clj#L179) and is not covered with tests.
[Here](https://github.com/ztellman/aleph/blob/be1ee04c130660ce69ee7c6603132b36be79ec1f/src/aleph/netty.clj#L283-L297) we have quite a tricky machinery with switching on/off `autoRead` setting on a channel. Would be nice to have tests coverage for this.
Both [server](https://github.com/ztellman/aleph/blob/be1ee04c130660ce69ee7c6603132b36be79ec1f/src/aleph/http/server.clj#L314-L316) and [client](https://github.com/ztellman/aleph/blob/be1ee04c130660ce69ee7c6603132b36be79ec1f/src/aleph/http/client.clj#L164-L173) handlers simply log all caught exceptions. In case something goes wrong in the pipeline, processing might be stuck. E.g. when a server responds unexpectedly, Netty throws...
In some cases, you need to be able to clean up all resources allocated by the HTTP client: close all open connections, shutdown appropriate threads etc. It's useful and desirable...
Right now Aleph exposes 3 out of 5 configuration options for `HttpServerCodec` [here](https://github.com/ztellman/aleph/blob/568252716ef75ee0b0dca1e25ab5a2ab3d28d726/src/aleph/http/server.clj#L425-L430): * `maxInitialLineLenght` * `maxHeaderSize` * `maxChunkSize` Not supported: * `validateHeaders` (set to `false` by default) * `initialBufferSize`...
Right now there's no obvious way to cancel request processing as we do not have access to the information about connection status in the handler. With [this change](https://github.com/ztellman/aleph/commit/ae208a675d1b15f2d86f8a7dc77294dc7aa11012) I can...
It's not an Aleph issue specifically... more like a flaw of async event-driven HTTP communication. Unfortunately with practical implications. Let's assume I'm using HTTP client and 1. The server responds...
Working with SNI-enabled hosts on the client was discussed [here](https://github.com/ztellman/aleph/issues/317). Right now it's possible only through tricky `pipeline-transform` machinery. Same goes for the server as you need to plug in...
When I clone aleph, I get an error: ```console $ git clone --config transfer.fsckObjects=true [email protected]:ztellman/aleph.git aleph1 Cloning into 'aleph1'... remote: Counting objects: 7831, done. error: object 31dfaf308f5ea00c04c0653ecbe8937fb0add324: zeroPaddedFilemode: contains zero-padded...
