Commit Graph

20 Commits

Author SHA1 Message Date
Michael Bell f1a6056953 Immediately close bad connections to prevent file exhaustion
osrm-routed does not immediately clean up a keep-alive connection
when the client closes it. Instead it waits for five seconds
of inactivity before removing.

Given a setup with low file limits and clients opening and
closing a lot of keep-alive connections, it's possible for
osrm-routed to run out of file descriptors whilst it waits for
the clean-up to trigger.

Furthermore, this causes the connection acceptor loop to exit.
Even after the old connections are cleaned up, new ones
will not be created. Any new requests will block until the
server is restarted.

This commit improves the situation by:

- Immediately closing connections on error. This includes EOF errors
indicating that the client has closed the connection. This releases
resources early (including the open file) and doesn't wait for the
timer.

- Log when the acceptor loop exits. Whilst this means the behaviour
can still occur for reasons other than too many open files,
we will at least have visibility of the cause and can investigate further.
2021-09-04 01:55:36 +01:00
Daniel Patterson 50d9632ed7 Upgrade formatting to clang-format 10 (#5895)
* Update formatting tools to clang-format-10

* Reformat using clang-format-10.0.09
2020-11-26 07:21:39 -08:00
Denis Chaplygin 746b2e94a0 Formatted the code 2019-09-10 15:50:13 +03:00
Denis Chaplygin 43f60bc69c We should check, is connections is still open and gracefully return on connection closed by remote. 2019-09-10 14:32:16 +03:00
Denis Chaplygin 145974bc92 clear input/output buffers during keepalive clean-up procedure.
Should fix #5545
2019-09-09 11:20:47 +03:00
Denis Chaplygin cd7229ba06 connection's reply needs to be reset before handling next request on same connection. 2019-08-28 17:22:07 +03:00
Denis Chaplygin 9efcab2108 Receive timeout should only be active for a second and
following requests on a keep-alive connections.
2019-08-23 12:21:54 +03:00
Denis Chaplygin a7b7d77e1f Formatted the code. 2019-08-23 12:21:54 +03:00
Denis Chaplygin 22550d078f Added timeout handling for keep-alive operations. 2019-08-23 12:21:54 +03:00
Denis Chaplygin a0582a3e68 Added keep-alive support to the http server. 2019-08-23 12:21:54 +03:00
Patrick Niklaus 6e16eab6ec Format with clang-format 3.8 2016-05-27 21:05:04 +02:00
Patrick Niklaus 1554916639 Include <iterator> for back_inserter 2016-04-12 18:11:08 +02:00
Patrick Niklaus f3e72623e9 Add viaroute suport for new API 2016-04-05 22:58:32 +02:00
karenzshea f5c12ec433 comments 2016-01-29 23:49:09 +01:00
karenzshea 262cdd7bc0 move tribool obj into request_parser 2016-01-29 23:49:09 +01:00
Daniel J. Hofmann e7d9e42a16 Fix race condition in osrm-routed HTTP server
We had a similar issue a few month ago. Stumbled upon this by accident.

It goes like this:

- `output_buffer` is function-local
- we pass it to async_write and leave the scope
- `output_buffers` goes out of scope
- bad things happen, sometimes

The fix is to, again, put it inside the connection that handles itself
via the enable_shared_from_this idiom.

References:

- http://www.boost.org/doc/libs/1_59_0/doc/html/boost_asio/reference/async_write/overload1.html
- https://github.com/Project-OSRM/osrm-backend/pull/1690
2016-01-13 23:21:45 +01:00
Patrick Niklaus 6b18e4f7e9 Add namespace around all files 2016-01-08 01:30:52 +01:00
Daniel J. Hofmann 685d330ee2 Remove license headers from project source files; global license 2016-01-04 19:06:32 +01:00
Patrick Niklaus 97585bb26a Fix include paths 2016-01-03 18:47:50 +01:00
Patrick Niklaus bfc6c9b89d Move files in src/ include/ 2016-01-03 16:37:38 +01:00