May 30, 2007
View 761bd3097e

Add two new config variables to set source IP on outbound TCP connections. As Jan stated this is useful for HA setups with virtual IPs.

Hendrik Scholz authored on 30/05/2007 12:48:24
February 20, 2007
View 895cb2af9e

- make tar doesn't exclude tls*, but instead tls/* - don't append -tls to the version number unless CORE_TLS is used (there's enough information in the flags to see if support for the tls modules is enabled)

Andrei Pelinescu-Onciul authored on 20/02/2007 19:41:14
February 10, 2007
View 6c53d41a09

- tls hooks support: special tls hooks added in core that allow a module or core part to provide a tls implementation. Support for the old in-core tls compile options is preserved (but one can compile with either tls hooks support or tls-in-core support, not with both in the same time). Changes were kept to a minimum. - Makefiles: make TLS=1 deprecated, use instead make TLS_HOOKS=1 or make CORE_TLS=1 (if in-core tls support is needed). Added TLS_EXTRA_LIBS (e.g. make CORE_TLS=1 TLS_EXTRA_LIBS="-lz -lkrb5").

Andrei Pelinescu-Onciul authored on 10/02/2007 18:52:48
January 27, 2007
View b8116f0486

- minor tcp fix: some bug detection code (which should never be triggered) is more robust now

Andrei Pelinescu-Onciul authored on 27/01/2007 08:01:07
November 4, 2006
View d8b11bbcd5

- tcp fixes: tcpconn_timeout(); expire now timeout; switched to "raw" ticks (the conversion to s was plagued by truncation errors => in some corner cases one could end up with tcp connections that never expire) - if tcp_con_lifetime is set to a negative value, use the maximum possible value instead

Andrei Pelinescu-Onciul authored on 04/11/2006 15:55:08
November 2, 2006
View c2518f5420

- more reliable tcp shutdown: force unlock the tcp lock in case some terminated process did hold it

Andrei Pelinescu-Onciul authored on 02/11/2006 23:35:32
View be7401cc34

- tcp fixes/atomic_t switch

Andrei Pelinescu-Onciul authored on 02/11/2006 22:55:56
October 2, 2006
View 144c394f42

- fork_process & fork_tcp_process fixes - reverted to the old fork()-in-parallel behaviour (uncomment FORK_DONT_WAIT for the "serial" fork()).

Andrei Pelinescu-Onciul authored on 02/10/2006 17:29:23
September 19, 2006
View 37209e1421

- applied patch from Dragos Vingarzan <vingarzan@fokus.fraunhofer.de> which moves all the forking part into 2 functions in pt.c (fork_process and fork_tcp_process). - added PROC_NOCHLDINIT rank value for Dragos's fork_process (if this value is used as child_id/rank_value the mod_child functions will not be called) - added register_procs(processes_no), used from mod_init when a module knows that it will fork some children (replaces the old process_count++ / the patch's estimated_process_count++) - added get_max_procs(): returns the maximum (estimated) number of processes

Andrei Pelinescu-Onciul authored on 19/09/2006 16:13:27
April 12, 2006
View e6a2b12e54

- msg_send(), udp_send() and tcp_send() parameter list changed (now they use a struct dest_info to pack several of the old params)

Andrei Pelinescu-Onciul authored on 12/04/2006 18:04:23
February 24, 2006
View 8f0e1af302

- tcp accounting: connections_no decreased on error - parse_uri: missing break for r2 - tm: enabled tm stats

Andrei Pelinescu-Onciul authored on 24/02/2006 18:35:35
February 22, 2006
View bf458eea46

- tcp_send fix: return <0 on send error. This is a modified version of Klaus Darilion patch/openser port (extra tcpconn_put on release error)

Andrei Pelinescu-Onciul authored on 22/02/2006 12:45:24
February 10, 2006
View e85eac0ab5

- removed some exported information: inactive_connection and number of requests (it was bogus anyway)

Andrei Pelinescu-Onciul authored on 10/02/2006 18:21:51
February 6, 2006
View d95154053d

- fixed some tcp connections accounting bugs - tcp info available via fifo/rpc

Andrei Pelinescu-Onciul authored on 06/02/2006 16:47:31
February 3, 2006
View 746f7674ae

- tcp children number is not anymore limited (was limited to 100) - tcp connection are passed to the tcp reader processes after some data is received on them and not immediately after accept(). This speeds up accepting lots of new connections. Old behaviour can be select by undef-ing TCP_PASS_NEW_CONNECTION_ON_DATA in tcp_main.c - lots of cleanups and small fixes - tsend_stream now support waiting forever (timeout==-1)

Andrei Pelinescu-Onciul authored on 03/02/2006 21:56:23