Hello,
We realized that redis_execute should return a negative value in case of failure (just like redis_cmd), instead of always returning 1. This introduced a problem with the behavior when calling redis_execute without giving a server name. Since this function will loop through all the servers and execute the pipelined commands for each one, in case of failure the user cannot determine for what server the error occurred, by only returning a negative value. This is why we decided to remove the possibility to call redis_execute with no arguments, and have it loop through all defined servers. The user should only call redis_execute(srv_name) and check the return value for this.
You can view, comment on, or merge this pull request online at:
https://github.com/kamailio/kamailio/pull/1112
-- Commit Summary --
* ndb_redis: remove redis_execute command with no arguments
* ndb_redis: make redis_execute() return negative value on error
-- File Changes --
M src/modules/ndb_redis/doc/ndb_redis_admin.xml (10)
M src/modules/ndb_redis/ndb_redis_mod.c (22)
M src/modules/ndb_redis/redis_client.c (20)
M src/modules/ndb_redis/redis_client.h (1)
-- Patch Links --
https://github.com/kamailio/kamailio/pull/1112.patchhttps://github.com/kamailio/kamailio/pull/1112.diff
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/1112
Backport of commits:
1) 7036c0ea947d6450e6a23c2e5b2428c834352de0
2) ff3cf17ff3566cb08cb0a6b9f7f7d893cd908956
3) 66595f39a4ae419b054bb433f2451a69155d580f
You can view, comment on, or merge this pull request online at:
https://github.com/kamailio/kamailio/pull/1113
-- Commit Summary --
* core: Added IP_FREEBIND flag durring socket initialization
-- File Changes --
M cfg.lex (2)
M cfg.y (3)
M globals.h (1)
M modules/sctp/sctp_server.c (15)
M receive.c (2)
M tcp_main.c (11)
M udp_server.c (10)
-- Patch Links --
https://github.com/kamailio/kamailio/pull/1113.patchhttps://github.com/kamailio/kamailio/pull/1113.diff
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/kamailio/kamailio/pull/1113
Why is all K related github messages also sent to sr-dev list? There is
lots of duplicate traffic because of it if sr-dev member is also
watching github notifications.
Or is it so that sr-dev members are not supposed to subscribe to
kamailio github notifications?
-- Juha
Hi,
I'm trying to fix tls.so to work with libressl. And there was already
one commit handling some of the api differences.
The next problem is that libressl ships the memory allocator changing
function, but it's a stub that just returns "not supported".
It seems that kamailio module does not load unless it can make libssl
use shm_alloc. I'm wondering if this is a hard requirement? Does the
tls module share libtls allocated things amongs forks? Or this purely
to track memory in kamailio framework?
If it's hard dependency is there any way to workaround this libressl
design decision?
I think one option to try is to just implement the CRYPTO_* allocation
function in kamailio main executable - they should then take preference
over the libcrypto ones. Thoughts?
Thanks,
Timo