sr-dev March 2015

sr-dev@lists.kamailio.org

34 participants
192 discussions

usrloc and db only mode
by Daniel-Constantin Mierla 21 Apr '15

21 Apr '15

Hello, following some discussions on github issues, mailing list and irc recently, I am opening a topic and see if we can get something to accommodate the needs out there at the best for everyone. These are more in the context with the increase of sizes for deployments, but also having the option to use a no-sql, some internal bits of usrloc may need tunings. The points to be considered: 1) discovering what is the server that wrote the record 2) efficiently getting the records that require nat keepalive The two are sometime related, but not necessary always. Victor Seva on irc channel was saying that is planning to propose a patch for 1), relying on socket value in the location table. The use case presented was many servers saving and looking up the records in the same table, but fetch only the records written by current instance for sending keepalives. It makes sense not send keepalives from all instances, for load as well and bandwidth considerations. This could work relying on local socket, but I am not sure how efficient will be when listening on many sockets (different ips, but also different ports or protocols - ipv4/6, tcp, tls, udp -- all add sockets). On 2), it's being for a while on my list to review, as the module uses custom sql queries (or even functions) to be able to match the the records for sending keepalives -- it does matching with bitwise operations to see if flags for nat are set. That doesn't work with no-sql databases (including db_text, db_mongodb and I expect db_cassandra). Even for sql, that kind of query is not efficient, especially when dealing with a large amount of records. The solutions that came in my mind for now: - for 1) -- add the server_id as a new column in the location table. It was pointed to be that ser location had it. This will turn a query matching on many socket-representation strings to one expression done on an integer. The value for server_id can be set via the core parameter with the same name. - for 2) -- add a new column 'keepalive' to be set internally by the module to 1 if any of the flags for sending keepalive was set. Then the query to fetch it will just use a match on it, rather that bitwise operations in the match expression. Furthermore, this can be set to a different value when enabling the keepalive partitioning (ie., sending keepalives from different timer processes, each doing for a part of natted users) - right now, for the db only mode, the selection is done by column id % (number of keepalive processes). Opinions, improvements, other proposals? Cheers, Daniel -- Daniel-Constantin Mierla http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda Kamailio World Conference, May 27-29, 2015 Berlin, Germany - http://www.kamailioworld.com

5 11

[kamailio] set defaults of fifo and ctl to /var/run/kamailio (#89)
by Victor Seva 16 Apr '15

16 Apr '15

Following #48 and #80 You can view, comment on, or merge this pull request online at: https://github.com/kamailio/kamailio/pull/89 -- Commit Summary -- * ctl: use RUN_DIR env to set DEFAULT_CTL_SOCKET * etc: set ctl "binrpc" to new /var/run/kamailio default value * utils/kamctl: change fifo default path to /var/run/kamailio * etc: change fifo default to /var/run/kamailio -- File Changes -- M etc/kamailio-basic.cfg (4) M etc/kamailio-oob.cfg (4) M etc/kamailio.cfg (4) M modules/ctl/ctl_defaults.h (4) M utils/kamctl/kamctl.fifo (2) M utils/kamctl/kamctlrc (4) -- Patch Links -- https://github.com/kamailio/kamailio/pull/89.patch https://github.com/kamailio/kamailio/pull/89.diff --- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/pull/89

2 3

TM possible deadlock
by Jason Penton 09 Apr '15

09 Apr '15

Hi All, I have been experiencing a deadlock when a timeout occurs on a t_relayed() INVITE. Going through the code I have noticed a possible chance of deadlock (without re-entrant enabled). Here is my thinking: t_should_relay_response() is called with REPLY_LOCK when the timer process fires on the fr_inv_timer (no response from the INVITE that was relayed, other than 100 provisional) and a 408 is generated. However, from within that function there are calls to run_failure_handlers() which in turn *could* try and lock the reply (viz. somebody having a t_reply() call in the cfg file - in failure route block). This would result in another lock on the same transaction's REPLY_LOCK.... Has anybody else experienced something like this? this is on master btw. Cheers Jason

2 16

Planning to release Kamailio v4.2.4
by Daniel-Constantin Mierla 02 Apr '15

02 Apr '15

Hello, I am considering to release a new maintenance version from latest stable branch - to be v4.2.4 - next week, most likely on Thursday, April 2. If there is anything important to get in and not discussed yet, bring the topic on development mailing list. Cheers, Daniel -- Daniel-Constantin Mierla http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda Kamailio World Conference, May 27-29, 2015 Berlin, Germany - http://www.kamailioworld.com

1 1

[kamailio] Implement method to selectively disable MSILO offline notification (#61)
by Anthony Messina 01 Apr '15

01 Apr '15

If Kamailio is globally configured to send offline notification replies using `modparam("msilo", "from_address", "sip:$rU@example.com")`, there is currently no way to disable the offline notification reply during script processing. For example, a scenario where you might want to store the original MESSAGE but not send the offline notification reply is when you are also using the IMC module. When `user(a)example.com` is part of an IMC chat, but goes offline for some reason, the MSILO module will store original MESSAGE, then generate the offline notification reply back to the IMC chat, which generates another MESSAGE with `user(a)example.com` as a recipient... This instantly leads to thousands of MESSAGE generations. I am thinking that it's a nice feature to have offline notification replies enabled when `modparam("msilo", "from_address", "sip:$rU@example.com")` is defined, but that the MSILO module could check the existence (nor non-existence) of a flag to determine whether or not it would generate an offline notification reply, so the logic would be something like: ``` #!define FLT_MSILO_DISABLE_OFFLINE_REPLY 13 modparam("msilo", "from_address", "sip:$rU@example.com") modparam("msilo", "disable_offline_reply_flag", FLT_MSILO_DISABLE_OFFLINE_REPLY) ``` Then m_store() checks that `from_address` is valid and that `disable_offline_reply_flag` is not set. --- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/61

2 2

[kamailio] db_sqlite: Core dump at mod_Init (#43)
by Olle E. Johansson 01 Apr '15

01 Apr '15

Trying to load my server in the background, the main process core dumps at init of db_sqlite Git head, os/x (gdb) bt full #0 0x00007fff8610c66a in _dispatch_barrier_async_f_slow () No symbol table info available. #1 0x00007fff8acd13bd in sqlite3_initialize () No symbol table info available. #2 0x0000000104b75ba3 in sqlite_mod_init () at db_sqlite.c:69 No locals. #3 0x00000001033bc93c in init_mod (m=0x103809b80) at sr_module.c:943 No locals. #4 0x00000001033bc414 in init_mod (m=0x103809e58) at sr_module.c:940 No locals. --- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/43

2 5

[kamailio] websocket: `tcpconn_get()` error gets turned into `E_OUT_OF_MEM` which becomes "500 No Error" (#47)
by Walter Doekes 01 Apr '15

01 Apr '15

My websocket TLS server is full of these kinds of messages: Jan 18 18:10:26 ws0 /usr/sbin/kamailio[19701]: NOTICE: <script>: http:217.120.x.x:55386: WS connection closed ... Jan 18 18:10:26 ws0 /usr/sbin/kamailio[19689]: WARNING: <core> [msg_translator.c:2506]: via_builder(): TCP/TLS connection (id: 0) for WebSocket could not be found Jan 18 18:10:26 ws0 /usr/sbin/kamailio[19689]: ERROR: <core> [msg_translator.c:1722]: build_req_buf_from_sip_req(): could not create Via header Jan 18 18:10:26 ws0 /usr/sbin/kamailio[19689]: ERROR: tm [t_fwd.c:527]: prepare_new_uac(): could not build request Jan 18 18:10:26 ws0 /usr/sbin/kamailio[19689]: ERROR: tm [t_fwd.c:1773]: t_forward_nonack(): ERROR: t_forward_nonack: failure to add branches Jan 18 18:10:26 ws0 /usr/sbin/kamailio[19689]: ERROR: sl [sl_funcs.c:387]: sl_reply_error(): ERROR: sl_reply_error used: No error (2/SL) (repeat these last errors for a bunch of attempted NOTIFY forwards) The route block does basically something like this: # add_contact_alias(); # only for requests from the outside loose_route(); if (!t_relay()) { sl_reply_error(); } The problem arises here: }else if (send_info->proto==PROTO_WS){ ... con = tcpconn_get(send_info->id, &ip, port, from, 0) ... if (con == NULL) { LM_WARN("TCP/TLS connection (id: %d) for WebSocket could not be found\n", send_info->id); The NULL failure status gets returned up to `prepare_new_uac` in `t_fwd.c`: shbuf=build_req_buf_from_sip_req( i_req, &len, dst, BUILD_IN_SHM); if (!shbuf) { LM_ERR("could not build request\n"); ret=E_OUT_OF_MEM; goto error01; } At this point, ser_error will become `E_OUT_OF_MEM` while it should be something like `E_SEND`. And `E_OUT_OF_MEM` gets translated to `500 No Error` because we're not running in DEBUG mode. What causes the connection to drop in the first place, you ask? 18:10:18.690738 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [S], seq 1323983240, win 8192, options [mss 1460,nop,wscale 8,nop,nop,sackOK], length 0 18:10:18.690863 IP 195.35.x.x.443 > 217.120.x.x.55386: Flags [S.], seq 4077761781, ack 1323983241, win 14600, options [mss 1460,nop,nop,sackOK,nop,wscale 3], length 0 18:10:18.710846 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [.], ack 1, win 256, length 0 18:10:18.808751 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [P.], seq 1:246, ack 1, win 256, length 245 ... 18:10:19.233415 IP 195.35.x.x.443 > 217.120.x.x.55386: Flags [.], ack 31348, win 5126, length 0 18:10:26.489764 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [P.], seq 31348:32473, ack 34578, win 255, length 1125 ... 18:10:26.501409 IP 195.35.x.x.443 > 217.120.x.x.55386: Flags [P.], seq 42255:42916, ack 46010, win 5046, length 661 18:10:26.527755 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [.], ack 36993, win 252, length 0 18:10:26.527860 IP 195.35.x.x.443 > 217.120.x.x.55386: Flags [.], seq 42916:47296, ack 46010, win 5278, length 4380 18:10:26.527888 IP 195.35.x.x.443 > 217.120.x.x.55386: Flags [FP.], seq 47296:48663, ack 46010, win 5278, length 1367 18:10:26.529179 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [.], ack 40501, win 254, length 0 18:10:26.529200 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [.], ack 42916, win 251, length 0 18:10:26.547276 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [.], ack 48664, win 251, length 0 18:10:26.549712 IP 217.120.x.x.55386 > 195.35.x.x.443: Flags [F.], seq 46010, ack 48664, win 251, length 0 18:10:26.549750 IP 195.35.x.x.443 > 217.120.x.x.55386: Flags [.], ack 46011, win 5278, length 0 Where you see that the FIN is initiated by `195.35.x.x` which is the Kamailio websocket server. The cause (probably) is the WS client closing the connection. In this case after re-subscribing with Expires:0. The presence server attempts to reply with a bunch of NOTIFYs with `Subscription-State: terminated;reason=timeout` but they bounce on the broken connection. If Kamailio would return a nice "477 Unfortunately error on sending to next hop occurred" it'd be prettier. Getting less "error" messages (a total of 6 *per* expired/unsubscribed subscription) after this error --which is apparently very common -- would be beneficial too. As for fixing: - We could change the `via_builder` to set `ser_error` (and check that in `build_req_buf_from_sip_req`), or - add error-code-out-parameters to all calls from `build_req_buf_from_sip_req` and down. I'm not sure if either is the best way. As for the excessive error reporting, would looking at `ser_error` before printing (another) error be an acceptable fix? Cheers, Walter Doekes OSSO B.V. --- Reply to this email directly or view it on GitHub: https://github.com/kamailio/kamailio/issues/47

2 4

git:master:845bdb88: presence: query_str filled with value of the corresponding column
by Daniel-Constantin Mierla 31 Mar '15

31 Mar '15

Module: kamailio Branch: master Commit: 845bdb8884743f5e126a53eb6aeb34f1f6afb1ea URL: https://github.com/kamailio/kamailio/commit/845bdb8884743f5e126a53eb6aeb34f… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2015-03-31T15:15:27+02:00 presence: query_str filled with value of the corresponding column - static value was used that made setting of parameter for column names not being considered --- Modified: modules/presence/presentity.c Modified: modules/presence/publish.c --- Diff: https://github.com/kamailio/kamailio/commit/845bdb8884743f5e126a53eb6aeb34f… Patch: https://github.com/kamailio/kamailio/commit/845bdb8884743f5e126a53eb6aeb34f… --- diff --git a/modules/presence/presentity.c b/modules/presence/presentity.c index c358b81..0f3c3e5 100644 --- a/modules/presence/presentity.c +++ b/modules/presence/presentity.c @@ -885,6 +885,7 @@ int pres_htable_restore(void) int event; event_t ev; char* sphere= NULL; + static str query_str; result_cols[user_col= n_result_cols++]= &str_username_col; result_cols[domain_col= n_result_cols++]= &str_domain_col; @@ -899,7 +900,7 @@ int pres_htable_restore(void) goto error; } - static str query_str = str_init("username"); + query_str = str_username_col; if (db_fetch_query(&pa_dbf, pres_fetch_rows, pa_db, 0, 0, 0, result_cols, 0, n_result_cols, &query_str, &result) < 0) { diff --git a/modules/presence/publish.c b/modules/presence/publish.c index 1b380bb..5a48f77 100644 --- a/modules/presence/publish.c +++ b/modules/presence/publish.c @@ -70,6 +70,7 @@ void msg_presentity_clean(unsigned int ticks,void *param) int i = 0, num_watchers = 0; presentity_t pres; str uri = {0, 0}, event, *rules_doc = NULL; + static str query_str; LM_DBG("cleaning expired presentity information\n"); if (pa_dbf.use_table(pa_db, &presentity_table) < 0) @@ -97,7 +98,7 @@ void msg_presentity_clean(unsigned int ticks,void *param) result_cols[etag_col=n_result_cols++] = &str_etag_col; result_cols[event_col=n_result_cols++] = &str_event_col; - static str query_str = str_init("username"); + query_str = str_username_col; if (db_fetch_query(&pa_dbf, pres_fetch_rows, pa_db, db_keys, db_ops, db_vals, result_cols, n_db_cols, n_result_cols, &query_str, &result) < 0)

1 0

git:master:47d506cd: presence: variable for last subscription uri made static and renamed
by Daniel-Constantin Mierla 31 Mar '15

31 Mar '15

Module: kamailio Branch: master Commit: 47d506cd6d6045bd09d89cafab0384d2a74a9c65 URL: https://github.com/kamailio/kamailio/commit/47d506cd6d6045bd09d89cafab0384d… Author: Daniel-Constantin Mierla <miconda(a)gmail.com> Committer: Daniel-Constantin Mierla <miconda(a)gmail.com> Date: 2015-03-31T15:08:23+02:00 presence: variable for last subscription uri made static and renamed - avoid eventual conflicts in the future because presence modules bind between each other --- Modified: modules/presence/subscribe.c --- Diff: https://github.com/kamailio/kamailio/commit/47d506cd6d6045bd09d89cafab0384d… Patch: https://github.com/kamailio/kamailio/commit/47d506cd6d6045bd09d89cafab0384d… --- diff --git a/modules/presence/subscribe.c b/modules/presence/subscribe.c index 7f826f9..84f16e5 100644 --- a/modules/presence/subscribe.c +++ b/modules/presence/subscribe.c @@ -755,7 +755,7 @@ void msg_watchers_clean(unsigned int ticks,void *param) LM_ERR("cleaning pending subscriptions\n"); } -char* last_presentity = NULL; +static char* _pres_subs_last_presentity = NULL; int pv_parse_subscription_name(pv_spec_p sp, str *in) { @@ -787,7 +787,8 @@ int pv_parse_subscription_name(pv_spec_p sp, str *in) int pv_get_subscription(struct sip_msg *msg, pv_param_t *param, pv_value_t *res) { if(param->pvn.u.isname.name.n==1) /* presentity */ - return last_presentity == NULL ? pv_get_null(msg, param, res) : pv_get_strzval(msg, param, res, last_presentity); + return (_pres_subs_last_presentity == NULL) ? pv_get_null(msg, param, res) + : pv_get_strzval(msg, param, res, _pres_subs_last_presentity); LM_ERR("unknown specifier\n"); return pv_get_null(msg, param, res); @@ -848,9 +849,9 @@ int handle_subscribe(struct sip_msg* msg, str watcher_user, str watcher_domain) str reply_str; int sent_reply= 0; - if(last_presentity) { - pkg_free(last_presentity); - last_presentity = NULL; + if(_pres_subs_last_presentity) { + pkg_free(_pres_subs_last_presentity); + _pres_subs_last_presentity = NULL; } /* ??? rename to avoid collisions with other symbols */ @@ -952,9 +953,10 @@ int handle_subscribe(struct sip_msg* msg, str watcher_user, str watcher_domain) } if(subs.pres_uri.len > 0 && subs.pres_uri.s) { - last_presentity= (char*)pkg_malloc((subs.pres_uri.len+1) * sizeof(char)); - strncpy(last_presentity, subs.pres_uri.s, subs.pres_uri.len); - last_presentity[subs.pres_uri.len] = '\0'; + _pres_subs_last_presentity = + (char*)pkg_malloc((subs.pres_uri.len+1) * sizeof(char)); + strncpy(_pres_subs_last_presentity, subs.pres_uri.s, subs.pres_uri.len); + _pres_subs_last_presentity[subs.pres_uri.len] = '\0'; } /* mark that the received event is a SUBSCRIBE message */

1 0

Proposed Change to DMQ Notification Address Resolution
by Robert Boisvert 30 Mar '15

30 Mar '15

DMQ Developers, Our team would like to setup a DMQ bus <http://kamailio.org/docs/modules/4.2.x/modules/dmq.html#idp2640048> that contains more than one notification address <http://kamailio.org/docs/modules/4.2.x/modules/dmq.html#dmq.p.notification_…> to support a high availability, fault-tolerant system where multiple servers are used to maintain the list of DMQ nodes. The failure of any one DMQ node or the startup order should not cause the DMQ bus to lose track of available nodes. As nodes go offline and online the DMQ bus should be updated using information from the active nodes. To do this we propose changing the way DMQ uses the notification address. Currently the notification address resolves a DNS name to a single IP address even if a SRV record or multiple A/AAAA records are available. The proposal is to use all the A/AAAA records or SRV targets returned from a DNS name query as notification addresses. Do you have any comments, concerns, suggestions, or recommendations regarding this proposal? Your input is definitely welcome. Thank you, Bob

2 5

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

sr-dev March 2015