| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
| |
| |
| |
| |
| |
| | |
With epoll(7), MAX_EVENTS controls how many events can be retrieved in one
epoll_wait call, while with poll(2) this MAX_WATCHED controls how many
sockets or other FDs can be watched at once.
|
| |
| |
| |
| |
| |
| |
| | |
The previous count would be invalid at this point.
Should be possible to math out how many items are left, but this is left
as future work.
|
| |
| |
| |
| |
| |
| |
| | |
Weak tables are said to have suboptimal performance, so we might as well
get replace it with an increased default LRU cache size.
Sorry about the 'and'
|
| |
| |
| |
| | |
Should be done here too.
|
| |
| |
| |
| |
| | |
Since datamanager can fall back to the old method of loading the whole
list, which wouldn't come with a :close method.
|
| |
| |
| |
| | |
Maybe we need some sort of lint for this?
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Otherwise requests with Connection: close would be stuck in the async
wait that starts after the handle_request() call.
Together with the new async debugging, this makes the async thread stay
in the set of waiting runners forever, where previously it would simply
be garbage collected.
|
| |
| |
| |
| |
| | |
This closes the two FDs that the random access list abstraction uses,
otherwise they are left to the garbage collector.
|
| |
| |
| |
| |
| |
| | |
Gives some access to node details which are otherwise hard to determine
if you only see the plain text summary, since it is shared based on the
pubsub#type setting (or payload xmlns).
|
| | |
|
| | |
|
| |
| |
| |
| | |
Should skip a syscall for each write when using epoll.
|
| |
| |
| |
| |
| |
| | |
Better performance under load maybe?
See b890ceb1c24f for previous increase
|
| | |
|
| |
| |
| |
| |
| |
| | |
This is purely for informational purposes, so it's possible to determine
externally whether a runner is using the default runner function (which
executes functions as work items) or a custom runner function.
|
| |
| |
| |
| | |
Mostly expected to be useful for debugging purposes.
|
| |
| |
| |
| | |
Knowing the state of the coroutine as well as the runner state can be helpful.
|
| |
| |
| |
| | |
This can be used for debugging and introspection.
|
| |
| |
| |
| |
| | |
Should have been part of f33887f925e1 to ensure it won't skip processing
timers at all when very busy.
|
| |
| |
| |
| | |
Updates for 65fb0d7a2312::59c3d775c7fa
|
| |
| |
| |
| |
| | |
This would indicate that a reverse proxy is used, which gets to be
responsible for that since it probably holds the actual cert.
|
| | |
|
| |
| |
| |
| |
| |
| | |
The 'anonymous_login' setting is deprecated and prosodyctl check config
will tell you to change it to 'authentication = "internal_hashed"', so
we shouldn't need to care about here anymore.
|
| |
| |
| |
| | |
This code is hard to follow and in need of some refactoring.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
./tools/build-env/build.sh
Creates a container image based on Debian or Ubuntu
./tools/build-env/here.sh
Starts a container and mounts in the current working directory, from
where one can ./configure; make; make test etc
|
| | |
|
| |
| |
| |
| |
| | |
Had a hard time following what was happening when it did not specify
which grant or token was being removed.
|
| |
| |
| |
| | |
Connection: keep-alive is implicit in HTTP/1.1 but explicit > implicit
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This should speed up repeated requests to the same site by keeping their
connections around and sending more requests on them.
Sending multiple requests at the same time is not supported, instead a
request started while another to the same authority is in progress would
open a new one and the first one to complete would go back in the pool.
This could be investigated in the future.
Some http servers limit the number of requests per connection and this
is not tested and could cause one request to fail, but hopefully it will
close the connection and prevent it from being reused.
|
| |
| |
| |
| |
| | |
Potentially allows sorting on those directly as they will be in
increasing order.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Allows sorting by id as a substitute for sorting by timestamp since it
has the timestamp in the encoded in the first part, and only things that
happen extremely close together may get out of order by such a sort,
which might not matter.
From draft-ietf-uuidrev-rfc4122bis formerly draft-peabody-dispatch-new-uuid-format
|
| |
| |
| |
| | |
modules_enabled (thanks aab and Menel)
|
| |
| |
| |
| | |
Security considerations added, no protocol changes.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| |
| | |
Previously the whole grant was deleted if it found one expired toke,
which was not indented.
|
| | |
|
| |
| |
| |
| | |
initialization
|
| |
| |
| |
| | |
Not the prosodyctl check dane I wanted to make but a start.
|
| |
| |
| |
| |
| |
| |
| | |
One small refactor but one huge step in the right direction
Mostly because adding another check would make the line checking for a
valid check exceed the column limit.
|
| |
| |
| |
| | |
Does this make it clearer what is going on?
|
| | |
|
| |
| |
| |
| | |
Simplifies configuration, only one already existing boolean to flip.
|