A filemap is a way of telling Kore to serve files from a directory
much like a traditional webserver can do.
Kore filemaps only handles files. Kore does not generate directory
indexes or deal with non-regular files.
The way files are sent to a client differs a bit per platform and
build options:
default:
- mmap() backed file transfer due to TLS.
NOTLS=1
- sendfile() under FreeBSD, macOS and Linux.
- mmap() backed file for OpenBSD.
The opened file descriptors/mmap'd regions are cached and reused when
appropriate. If a file is no longer in use it will be closed and evicted
from the cache after 30 seconds.
New API's are available allowing developers to use these facilities via:
void net_send_fileref(struct connection *, struct kore_fileref *);
void http_response_fileref(struct http_request *, struct kore_fileref *);
Kore will attempt to match media types based on file extensions. A few
default types are built-in. Others can be added via the new "http_media_type"
configuration directive.
The HTTP layer used to make a copy of each incoming header and its
value for a request. Stop doing that and make HTTP headers zero-copy
all across the board.
This change comes with some api function changes, notably the
http_request_header() function which now takes a const char ** rather
than a char ** out pointer.
This commit also constifies several members of http_request, beware.
Additional rework how the worker processes deal with the accept lock.
Before:
if a worker held the accept lock and it accepted a new connection
it would release the lock for others and back off for 500ms before
attempting to grab the lock again.
This approach worked but under high load this starts becoming obvious.
Now:
- workers not holding the accept lock and not having any connections
will wait less long before returning from kore_platform_event_wait().
- workers not holding the accept lock will no longer blindly wait
an arbitrary amount in kore_platform_event_wait() but will look
at how long until the next lock grab is and base their timeout
on that.
- if a worker its next_lock timeout is up and failed to grab the
lock it will try again in half the time again.
- the worker process holding the lock will when releasing the lock
double check if it still has space for newer connections, if it does
it will keep the lock until it is full. This prevents the lock from
bouncing between several non busy worker processes all the time.
Additional fixes:
- Reduce the number of times we check the timeout list, only do it twice
per second rather then every event tick.
- Fix solo worker count for TLS (we actually hold two processes, not one).
- Make sure we don't accidentally miscalculate the idle time causing new
connections under heavy load to instantly drop.
- Swap from gettimeofday() to clock_gettime() now that MacOS caught up.
- Change pools to use mmap() for allocating regions.
- Change kore_malloc() to use pools for commonly sized objects.
(split into multiple of 2 buckets, starting at 8 bytes up to 8192).
- Rename kore_mem_free() to kore_free().
The preallocated pools will hold up to 128K of elements per block size.
In case a larger object is to be allocated kore_malloc() will use
malloc() instead.
This basically turns off the HTTP layer for Kore. It does not
compile in anything for HTTP.
This allows Kore to be used as a network application platform as well.
Added an example for this called nohttp.
Other changes that sneaked in while hacking on this:
* Use calloc(), kill pendantic malloc option.
* Killed off SPDY/3.1 support completely, will be superseded by http2
Note that comes with massive changes to a lot of the core API
functions provided by Kore, these might break your application.
Change the callback prototypes to:
void callback(struct kore_msg *msg, const void *data);
This allows the callbacks to receive the full kore_msg data structure
as sent over the wire (including length and id). Useful for future
additions to the kore_msg structure (such as worker origin).
Several other improvements:
* Accesslog now uses the msg framework as well.
* Websocket WEBSOCKET_BROADCAST_GLOBAL now works.
Small websocket improvement in this commit:
* Build the frame to be sent only once when broadcasting
instead of per connection we are broadcasting towards.
Introduces a few new api functions:
- kore_websocket_handshake(struct http_request *):
Performs the handshake on an HTTP request (coming from page handler)
- kore_websocket_send(struct connection *, u_int8_t, void *, size_t):
Sends data to a websocket connection.
- kore_websocket_broadcast(struct connection *, u_int8_t, void *, size_t, int):
Broadcast the given websocket op and data to all connected
websocket clients on the worker. Note that as of right now
the WEBSOCKET_BROADCAST_GLOBAL scope option does not work
yet and messages broadcasted will be restricted to workers
only.
- kore_worker_websocket_broadcast(struct connection *, void *, void *):
Backend function used by kore_websocket_broadcast().
Could prove useful for developers to have access to.
A simple example is given under examples/websocket.
Known issues:
Kore does not support PING or CONT frames just yet.
- The net code no longer has a recv_queue, instead reuse same recv buffer.
- Introduce net_recv_reset() to reset the recv buffer when needed.
- Have the workers spread the load better between them by slightly
delaying their next accept lock and giving them an accept treshold
so they don't go ahead and keep accepting connections if they end
up winning the race constantly between the workers.
- The kore_worker_acceptlock_release() is no longer available.
- Prepopulate the HTTP server response header that is added to each
response in both normal HTTP and SPDY modes.
- The path and host members of http_request are now allocated on the heap.
These changes overall result better performance on a multicore machine,
especially the worker load changes shine through.
This way we can get our code called whenever a stream is
completed. This cb handler does stand alone from an http_request
and is passed a netbuf data structure.
* Always make sure we end the stream properly
* Check for SPDY_FLOW_WINDOW_MAX on window frame updates
* Kill SPDY_STREAM_BLOCKING, once flow control kicks in its per session
Personally use this for testing Kore its performance without
letting the OpenSSL stack get in the way too much.
Note that it leaves data structures as is, and just removes
any calls to OpenSSL (and removes the linking vs OpenSSL).
- Attempt to chain as much as we can in the send netbufs
(keeps down the SSL_write calls, silly seeing it go out with 8 bytes)
- Change NETBUF_SEND_PAYLOAD_MAX to 4K
- Call SSL_write() with max NETBUF_SEND_PAYLOAD_MAX in size
- Go back to flushing the send buffers after each request
- No more need for a fixed pool for nb->buf, go back to the heap for now
- Disable Nagle, we're doing the chaining now anyway
No longer takes callbacks, flags, or *out arguments.
Update rest of the code that called these callbacks whenever sending
was completed, instead call them right away now.
Joris Vink
Stig Telfer
Ansen Dong
Geenz
Thordur Bjornsson