The redis onekey concepts is introduced to reduce traffic to redis
and redis notification traffic.
It modifies the current structure for one call in redis, which are
multiple keys with pre- and postfixes and the callid in between to
one key with the structure "json-<callid>". The value is a json
formatted string with the previous multi-key identifiers in it.
This flag can be used by the signaling proxy in order to tell rtpengine not to persist the call into Redis upon receiving offer/answer() control commands
Updating, freeing, and writing the recording metadata is now
threadsafe. This is in regards to the metadata that we receive from
the `rtpengine_offer` or `rtpengine_answer` commands, not the
`meta_fp` file.
We can simply use the `call->master_lock` variable for protecting
updates to the recording metadata. We had to move the metadata
handling code in call_interfaces.c up into the master_lock guarded
section. The code in "recording.c:meta_finish_file" is called with
master_lock protection already.
We write out what RTP packet number the SDP came before. When we receive
an SDP offer or answer, we write out the RTP packet number that the SDP
preceeded. This will let us split the RTP recording PCAP around SDP
renegotiation for things like hold/unhold.
Also write out whether the SDP was an offer or an answer.
We want to be able to match up the source and target <IP>:<PORT> with
the IP and port info sent in SDP. I adjusted the source and target
<IP>:<PORT> we write to the IP and UDP headers in the PCAP file.
I added the offer SDP (in addition to the already written answer SDP)
to the metadata file.
As a consequence of writing out the offer SDP, we will have empty PCAP
files and associated metadata files for calls that are never answered.
squash! Fixes to match PCAP packets with SDP
Added comments detailing the types of endpoint IP/port values we can
use.
We had initialization code for recording scattered through
"call_interfaces.c", "call.c", and "recording.c". I moved more of the
actual code into functions within recording.c under the parent function
`detect_setup_recording`. We call this function from "call_interfaces.c".
I moved the disjointed bit of PCAP initialization to occur right below
where we toggle recording on or off.
We record the answer SDP (not offer) to the metadata file, separeted by
newlines. It is after the PCAP file URL and before the timestamps. Maybe
we should be writing the rewritten SDP.
I refactored the handling of the "record-call" flag in the
"rtpengine_offer" and "rtpengine_answer" handler. We now set the
recording data structures in a function called `set_record_call`.
We also only handle "record-call" flags on OP_ANSWER for SDP answers.
When we receive an offer or an answer, we will only turn recording on if
it contains the "record-call=yes" flag. Likewise, we only turn recording
off if it contains the "record-call=no" flag. If no flag is specified,
the call keeps its current recording status. It used to be the case that
not specifying "record-call=yes" would turn off call recording.
This fix is necessary to account for SDP renegotiation during
hold/unhold. If the call system sends over another offer during SDP
renegotiation to hold and then unhold, we don't want to change recording
behavior unless the call system specifically says so.
We want to be able to associate call files with a call without the
presence of identifying metadata within the metadata file. To accomplish
this, we prepend the call-id to the start of the pcap recording files
and the call metadata files.
Even though call-id is supposed to be unique, because of paranoia we
keep some of the random affix hex string, but we reduced it down to an
8-byte random value.
Also, some minor argument ordering and name refactoring for random
string generation functions.
We used to sometimes free the generic metadata (passed in through
rtpengine commands) before writing it to disk. Then we were writing
blank metadata to our metadata files. We fixed the ordering of
our `free` operations.
This involved moving all code from fs.(c|h) to recording.(c|h).
We still spoof packets, so the UDP will look like all monologues are coming
over the same port and will probably look like they are all one stream if
you look at the PCAP file.
Command line option is "--recording-dir".
Renamed inner recording spool "recordings" to "pcaps".
This is to avoid name sharing conflicts with the "--recording-dir" command
line option, which specifies the recordings spool directory, and the
"$RECORDING_DIR/recordings" inner directory. Changing the inner directory
name to "pcaps" removes this name collision.
In the process, I changed the function names in fs.h to be consistent with
other functions. The names are structure like "$OBJECT_$VERB".
Fixes and changes:
- Only create the metadata file if the call is being recorded.
- Only write to the metadata file if we actually created it (NULL check).
- Make sure we have metadata before putting it on the call object
- Correctly overwrite recording metadata without leaking memory
- Set the no kernalization flag per call instead of for *every* packet.
- Logging cleanup.
Pass in "record-call" flag over `rtpengine_offer` or `rtpengine_answer`
message. RTP Engine tracks files used to record pcaps and send them back
in the response message.
Pipes call audio (unencrypted from both ends) to recording files.
Sets up file descriptors for local files to dump RTP recordings.
A file and a file descriptor per monologue in a call.
Recorded streams will be running in user daemon mode, not in kernel mode.
This removes first 12 octets from packet to record just the rtp.
- add --subscribe_keyspace list config parameter.
- don't delete foreign calls by timers
- fix synchronization of foreign calls (use a separate redis_notify database)
- fix statistics for control channel calls.
- fix deletion of foreign calls upon del notifications
- update rtpengine-ctl tool
Removes the explicit redis-read-db configuration and reduces the option
to one redis DB and one redis write DB. If only the redis DB is
configured, then it will be used for all operations. If both are
configured, then the redis DB will be used for reading and the write DB
will be used for writing (updates).
Change-Id: I8d5a32c53c9416b514c98d69c3afe7c547e530ad
The session limit is only for calls an rtpengine is responsible for.
Foreign calls (coming in via redis notification) are not counted as
long as the rtpengine is not responsible for those calls.
At least that means that the limit may exceed if the calls the rtpengine
is responsible for plus the former foreign calls are greater than the limit.
This will happen suddenly when the rtpengine becomes responsible for the
foreign calls.
Currently, every rtpengine will subscribe to redis-keyspace notification
so it will receive a notification when an call is inserted. If the call
is not already handeled by the rtpengine, the call will be restored.
The reason for this is to have in-place redundancy. Imagine you have
multiple rtpengines running, eachone will have all calls of the others.
When one rtpengine fails somehow, infrastructure guys use BGP in order to
'move' the IP address from one rtpengine to another. Thisone can handle
the new calls instantly since they're already recovered by
redis-notification feature.
Next step is internally identify those calls in order to prevent some
timers to delete the calls where no RTP flows. Second will be
something we call 'partitioning'. It means that the subscription
to a redis notify will only be for the keyspace a dedicated rtpengine
writes to. This leads to the point that you can make redundancy groups
(partitions) of the rtpengines.
This brings master up to date with branch `rfuchs/socket-rework` at
commit `b1bcc096b7`. The branches have diverged too much for a proper
merge, so this is a manual (squashed) merge.
The old master before this merge can be found in branch
`old-master-before-socket-rework` (commit `82199216b2`).
This is a complete rewrite of all socket handling routines. The most
important functional change is that sockets aren't indiscriminately
bound to INADDR_ANY (or rather in6addr_any), but instead are always
bound to their respective local interface address and with the correct
address family.
Side effects of this are that in multi-homed environments, multiple
sockets must be opened (one per interface address and family) which must
be taken into account when considering RLIMIT_NOFILE values. As a
benefit, this change allows rtpengine to utilize the full UDP port space
per interface address, instead of just one port space per machine.
The socket abstraction also makes it possible to support RTP over TCP in
the future.
Change-Id: If6cf4f42136229490186d2d2482fb4fc140c2b53
* 0 will allow 0 sessions (e.g. can be used for draining rtpengine)
* -1 will disable the feature and will be treated the same way as if
MAX_SESSIONS variable has not been set via configuration file
(or has been set to -1 in configuration file)
* < -1 will not be taken into consideration
* add check for minint range also
This enables switching over the endpoint of a dialogue to a new client
with potentially different capabilities
Change-Id: I973b3c5d004014cc73e6ebc3e87d03c742bc1951
(cherry picked from commit 69b7586aa4)
(cherry picked from commit 93332f0f6e)
use via-branches to predict different destinations for branched SDP
offers without knowing the respective to-tag ahead of time.
Squashed commit of the following:
commit 0e81dc9828
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu May 21 11:06:15 2015 -0400
fix segfault due to missing reverse tagging
commit cd7a26314b
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu May 21 08:57:42 2015 -0400
support branched offer with previously unseen to-tag
commit 77da616dd2
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Apr 29 11:15:37 2015 -0400
implement via-branch handling
Squashed commit of the following:
commit 00213e66c7
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 27 09:40:04 2015 -0500
perform ICE restart if we change ports
commit 27fbcbd6cb
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 15:58:12 2015 -0500
locking fixes
commit 60c1c5ae13
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 15:58:06 2015 -0500
unkernelize when ICE completes
commit 1d816f9864
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 15:54:44 2015 -0500
relaxed locking where not needed
commit 75b58a9093
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 11:46:08 2015 -0500
use atomic bitfield instead of bitfield+mutex
commit 03552eeed9
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 11:24:59 2015 -0500
shuffle around aux.h a bit more
commit b9b8a3aa5e
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 11:16:12 2015 -0500
remove some code redundancy through a "state machine"
commit 0b4bfef1b1
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 10:10:41 2015 -0500
reorder to match struct
commit a2a51d81a8
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 10:10:03 2015 -0500
clear more states for ICE restart
commit d554a2b858
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 09:58:29 2015 -0500
dont duplicate candidates on re-invite
commit 4c804652b7
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 09:58:20 2015 -0500
retain ICE role across re-invites
commit 4a586dd72d
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 09:42:09 2015 -0500
eliminate duplicate log messages
commit ef0be2e308
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 26 09:32:36 2015 -0500
fix incorrect log message
commit 2544b60f00
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 15:26:48 2015 -0500
better logging for ICE agents
commit c42807384e
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 15:18:27 2015 -0500
dont run ICE checks if we dont have a password
commit 1b56cb75b5
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 15:16:19 2015 -0500
ICE pairs should go in triggered queue only once
commit d10c56f3ae
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 14:04:00 2015 -0500
obsolete the ICE agent running flag
commit 52237e3399
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 14:02:06 2015 -0500
cease checks once ICE is completed
commit 5332d18612
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 13:53:57 2015 -0500
fix ICE completion logging
commit 85f5fd63aa
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 13:53:33 2015 -0500
make better use of bit flags
commit 796b48bb78
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 12:09:46 2015 -0500
improved learned candidate pairing and completion logic
commit d15561072e
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 11:21:45 2015 -0500
support upper case transport strings
commit 557da7b1c3
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 10:43:57 2015 -0500
use struct endpoint in ice candidates
commit 951040bfd6
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 10:31:13 2015 -0500
more meaningful ICE log messages
commit 8ec2426bd3
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 09:54:49 2015 -0500
shut down agent if no components - limit number of candidates
commit 149260f3a6
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 25 09:44:13 2015 -0500
handle ICE restarts
commit 6a18c31f81
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 16:29:01 2015 -0500
dont clear succeeded flag when nominating
commit 93e0861d02
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 16:21:40 2015 -0500
use correct pwd in stun binding response
commit 32ba3ea406
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 16:03:09 2015 -0500
use deterministic foundation for prflx cands
commit 2f2dc91515
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 15:21:13 2015 -0500
handle initial ICE role
commit a6b8ad25e6
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 14:43:47 2015 -0500
another ICE scheduling fix
commit c572b04e55
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 14:14:29 2015 -0500
make ICE aware of rtcp-mux
commit 93cd2d2560
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 14:13:27 2015 -0500
print timestamp when logging to stderr
commit 22a52ffda2
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 14:13:13 2015 -0500
ICE scheduling fixes
commit 5d2d1a7739
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 13:27:59 2015 -0500
increase ICE pwd length to make chrome happy
commit ceff6698db
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 13:09:54 2015 -0500
dont discard RTP if ICE hasnt finished yet
commit e809877d0e
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 13:06:31 2015 -0500
implement remote-candidates (untested)
commit 41670eadbb
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 11:58:13 2015 -0500
shut down ICE agent when everything fails
commit 1ca26c4a81
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 11:10:46 2015 -0500
fix up SDP output for ICE candidates
commit 0287d68f33
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 10:09:49 2015 -0500
process ICE completion and fire up DTLS if desired
commit 5b6386036b
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 09:37:52 2015 -0500
use a btree to schedule ice checks
commit 2bc25f1e0f
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 24 09:37:02 2015 -0500
convert shutdown condition into global var
commit ecf0c5587c
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Mon Feb 23 14:50:46 2015 -0500
replace poller_now with timeval g_now
commit 164ecdd7ac
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Mon Feb 23 13:00:15 2015 -0500
handle nominations if we're controlling
commit d013659365
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Mon Feb 23 12:11:40 2015 -0500
copy controlling/ed role into agent
commit 09f1cae14a
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Mon Feb 23 12:02:54 2015 -0500
separate nominated pairs from valid pairs
commit f75f338cde
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Mon Feb 23 11:55:30 2015 -0500
organize aux.h into sections
commit d6acee1392
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Mon Feb 23 11:55:17 2015 -0500
use b-tree for various ICE lists
commit af9804d139
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 16:21:09 2015 -0500
prepare to finalize ice processing
commit 18df118375
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 16:19:50 2015 -0500
prettier logging
commit 280755c61a
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 15:02:14 2015 -0500
implement unfreezing of other components on success
commit 5d13657d5b
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 15:00:24 2015 -0500
bitfield access macros
commit 71746ad6a1
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 14:15:05 2015 -0500
handle ice updates and duplicate candidates
commit 02309d1b5b
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 12:22:29 2015 -0500
handle role conflicts
commit 52acf54ba5
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 10:24:01 2015 -0500
proper pair priorities calculation
commit 307af79e8d
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 09:54:18 2015 -0500
fix address family mixups
commit 7cbfd4d36a
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 20 09:14:49 2015 -0500
delay dtls startup and timeout checks while ice is running
commit 2a8ab75228
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 19 16:47:56 2015 -0500
process ice/stun responses
commit 92da323dcf
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 19 15:14:30 2015 -0500
adding stun server script for testing
commit c5cfeb122c
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 19 13:42:40 2015 -0500
act on stun requests, learn prflx candidates
commit 1cafd35e7a
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 19 13:30:14 2015 -0500
fix pktinfo for ipv4 packets
commit 8e338b8426
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 19 11:48:55 2015 -0500
rework interface handling to prepare to learn ICE candidate
commit 09e365c142
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 18 16:24:42 2015 -0500
add some locking
commit 8fc7b75095
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 18 16:05:20 2015 -0500
extend logic in response to stun req and implement triggered checks
commit 35eeb04376
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 18 12:46:42 2015 -0500
handle ice/stun retransmits and timeouts
commit b5637565b6
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 18 11:19:19 2015 -0500
first implementation of sending ICE checks
commit f0c1928c05
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 17 14:39:11 2015 -0500
preliminary list and loop for ICE checks
commit c38d6e22c1
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Tue Feb 17 12:00:24 2015 -0500
pair up candidates and prepare to run checks
commit d9559b4c59
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 13 15:36:29 2015 -0500
parse and remeber basic ICE attributes
Squashed commit of the following:
commit cca40e8e31
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Feb 12 10:17:51 2015 -0500
support per-payload-type stats in kernel module
commit dcc0dc0002
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 11 12:14:44 2015 -0500
count unknown rtp type as error
commit 941bde0df5
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Wed Feb 11 12:01:55 2015 -0500
use the list of rtp formats from the m= line
This avoids dynamically altering the rtpstats hash table and makes
keeping packet stats lock free.
commit 9150fed671
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 6 15:06:04 2015 -0500
fix and simplify rtpmap hash table
commit 1f73741cbf
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Fri Feb 6 13:56:07 2015 -0500
MT#9935 understand, learn, remember and report RTP payload types (codecs)
commit b0d690837c
Author: Richard Fuchs <rfuchs@sipwise.com>
Date: Thu Nov 20 13:36:25 2014 -0500
parse a=rtpmap attribute
Claudiu Boriga
Frederic-Philippe Metz
Richard Fuchs
Pawel Kuzak
Dylan Mikus
Lucian Balaceanu
Eric Green
smititelu
Pawel Kuzak