sr-dev October 2019

sr-dev@lists.kamailio.org

22 participants
436 discussions

[kamailio/kamailio] mod pua (dialoginfo) etag race conditions (#2048)

by Laszlo

I'm observing the following scenario: * mod_dialog callbacks trigger 2 or more times (nearly) simultaneously for the same dialog * pua_dialoginfo sends PUBLISH 1, referencing etag A * pua_dialoginfo sends PUBLISH 2, referencing etag A * presence_dialoginfo processes PUBLISH 1, replies with new etag B * presence_dialoginfo processes PUBLISH 2, replies with a 412 (because etag A no longer exists) * pua_dialoginfo receives the 412 and re-tries it as PUBLISH 3 ("sent a PUBLISH within a dialog that no longer exists, send again an intial PUBLISH") * presence_dialoginfo processes PUBLISH 3, and may or may not accept it The situation as described is not ideal since it'll fill up your logs with errors, but isn't critical per se. Much more problematic is when there are more than 2 PUBLISHes generated for the same dialog simultaneously, as this can cause a (near) infinite race between the various PUBLISH requests all fighting to update the same etag. For example, 10 PUBLISH are sent out for etag A; all but one are rejected with a 412; then the other 9 keep on bouncing back and forth between pua_dialoginfo and presence_dialoginfo because they do not share the same view on the dialog's latest etag. Even worse is when presence_dialoginfo is rejecting *all* incoming PUBLISHes with a 412, for example because of a database/memory/replication problem or a malformed request. A `t_reply("412", "Not today")` in the presence_dialoginfo server, combined with a single PUBLISH from pua_dialoginfo is enough to reproducibly brick the pua_dialoginfo server because it runs into critical memory fragmentation levels. I think there are multiple ways to fix or alleviate this problem. ## pua generic * pua (publ_cback_func) should not retry 412-failed PUBLISHes indefinitely, but e.g. at most once * pua should not generate simultaneous PUBLISHes for the same presentity. It should delay PUBLISH 2 until PUBLISH 1 is either (permanently) accepted or rejected; or it should discard PUBLISH 2 immediately when it is generated. * Perhaps make handling of 412 replies more fine-grained. Currently every 412 reply is handled like this ("sent a PUBLISH within a dialog that no longer exists"), while that statement doesn't apply to all possible 412 replies. ## pua_dialoginfo specific * pua_dialoginfo currently subscribes to a lot of mod_dialog callbacks. For example subscribing to both DLGCB_CONFIRMED and DLGCB_CONFIRMED_NA will always get you two (rapidly succeeding) PUBLISHes with exactly the same contents. Subscribing to DLGCB_REQ_WITHIN means you'll get a new PUBLISH for every re-INVITE such as hold/unhold or codec negotiation, which is useless in many usecases. It would be helpful to allow configuring pua_dialoginfo with a list of callbacks to subscribe to. * With or without a smaller set of mod_dialog callbacks, pua_dialoginfo can generate multiple PUBLISH requests with exactly the same contents. Since pua is aware of the (last) state that it published for the presentity, it could compare if the newly generated PUBLISH is any different from the last known state, and discard it if it's not. -- 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/issues/2048

1 year

[kamailio/kamailio] kamctl - shell interpreter selection (#2019)

by Daniel-Constantin Mierla

### Description The `kamctl` tool seems to require `bash`, while trying to use `/bin/sh`, which can point to `dash` or other shell interpreters. For example, the output of `kamdbctl create`: ``` MySQL password for root: -e \E[37;33mINFO: creating database kamailio_simple_db ... -e \E[37;33mINFO: granting privileges to database kamailio_simple_db ... -e \E[37;33mINFO: creating standard tables into kamailio_simple_db ... -e \E[37;33mINFO: Core Kamailio tables succesfully created. Install presence related tables? (y/n): n /usr/sbin/kamdbctl: 216: /usr/sbin/kamdbctl: Bad substitution ``` It seems that the issue is expanding the variable when getting the answer for y/n question: * https://github.com/kamailio/kamailio/blob/master/utils/kamctl/kamdbctl.base… Such expression seems to be specific for bash: * https://mywiki.wooledge.org/Bashism ### Troubleshooting #### Reproduction Run `kamctl` with `/bin/sh` pointing to `bash`. ### Possible Solutions Decide what to do to have an acceptable solution: enforce `bash`, remove bashisms`... Or maybe focus to make `kamcli` a (full) replacement for `kamctl/kamdbctl` and get rid of those old-style shell/bash scripts: * https://github.com/kamailio/kamcli -- 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/issues/2019

1 year

[kamailio/kamailio] EVAPI: Better ability to remove dead clients (#1880)

by weili-jiang

### Description If an EVAPI client drops dead and silently closes its TCP connection, EVAPI is not notified in an acceptable time frame. The default OS TCP keep-alive parameters are generally far too long to be effective, and there is no way to override these using module or runtime configuration. Moreover, any attempts to put in a workaround is thwarted by the inability to close an arbitrary EVAPI connection from say, a timer event (since there is no way to specify a connection to close outside the EVAPI context). ### Expected behavior When an EVAPI client goes dead, the connection should be closed in the order of seconds (ideally configurable). The EVAPI should allow connections to be closed via scripting from any context. #### Actual observed behavior When an EVAPI client goes dead, the connection stays open (counting towards the maximum number of allowed clients). When the evapi_close is invoked from other contexts, nothing happens. #### Debugging Data #### Log Messages #### SIP Traffic ### Possible Solutions ### Additional Information * **Kamailio Version** - output of `kamailio -v` ``` version: kamailio 5.2.1 (x86_64/linux) flags: STATS: Off, USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC, Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLACKLIST, HAVE_RESOLV_RES ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144 MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: unknown compiled with gcc 6.3.0 ``` * **Operating System**: Debian GNU/Linux 9 (stretch) ``` Linux 3b83b180f7c9 4.15.0-45-generic #48-Ubuntu SMP Tue Jan 29 16:28:13 UTC 2019 x86_64 GNU/Linux ``` -- 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/issues/1880

1 year

[kamailio/kamailio] support for a kamailio-dev package for external modules (#1752)

by Victor Seva

>From https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=902944 > Hi! > > We've got several local kamailio modules, and it would be nice to be > able to build them separately and not have to hook them into the > kamailio source and build systems. I suppose others might find this > useful too. > > I started pondering what adding support for this would imply, and I > think the first step is whether this is something you think upstream > and you'd be happy supporting? It could be that the internal interfaces > are completely unstable and providing this might be a support nightmare, > for example, or other similar concerns. > > I think what would be needed is: > > - Support installing the library .so symlinks into the -dev package. > (The module .so I think would need to stay in the main and module > packages, because some stuff loads them as module.so directly?) > - Support installing the kamailio core, library and module .h files > into the -dev package. AFAIUI modules can have inter-module > dependencies and they might use interfaces from both kamailio core > itself, the shared libraries or other modules. > - Provide a pkg-config file with the necessary compile and link runes > (-I, -L, -rpath and similar) to be able to build 3rd party modules > easily. > > While checking this, I noticed there's a src/lib/Makefile.defs, which > has a TYPE variable to install headers and similar, but it does not > appear to be currently used? > > Thanks, > Guillem -- 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/issues/1752

1 year

[kamailio/kamailio] TLS: Consider what's new and what's supported in OpenSSL (#1878)

by Olle E. Johansson

### Description  ### Expected behavior #### Actual observed behavior #### Debugging Data ``` (paste your debugging data here) ``` #### Log Messages  ``` (paste your log messages here) ``` #### SIP Traffic  ``` (paste your sip traffic here) ``` ### Possible Solutions  ### Additional Information * **Kamailio Version** - output of `kamailio -v` ``` (paste your output here) ``` * **Operating System**:  ``` (paste your output here) ``` -- 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/issues/1878

1 year

[kamailio/kamailio] OpenSUSE warning cleanup (#1714)

by sergey-safarov

### Description During kamailio packaging on OpenSUSE in log exist this warnings ``` [ 860s] kamailio.i586: I: binary-or-shlib-calls-gethostbyname /usr/lib/kamailio/modules/acc_diameter.so [ 860s] kamailio.i586: I: binary-or-shlib-calls-gethostbyname /usr/lib/kamailio/modules/auth_diameter.so [ 860s] kamailio-json.i586: I: binary-or-shlib-calls-gethostbyname /usr/lib/kamailio/modules/jsonrpcc.so [ 860s] kamailio.i586: I: binary-or-shlib-calls-gethostbyname /usr/lib/kamailio/modules/pdb.so [ 860s] kamailio.i586: I: binary-or-shlib-calls-gethostbyname /usr/lib/kamailio/modules/siputils.so [ 860s] kamailio-xmpp.i586: I: binary-or-shlib-calls-gethostbyname /usr/lib/kamailio/modules/xmpp.so [ 860s] kamailio.i586: I: binary-or-shlib-calls-gethostbyname /usr/sbin/kamailio [ 860s] kamailio.i586: I: binary-or-shlib-calls-gethostbyname /usr/sbin/kamcmd [ 860s] The binary calls gethostbyname(). Please port the code to use getaddrinfo(). ``` Could you `port the code to use getaddrinfo()`. -- 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/issues/1714

1 year

[kamailio/kamailio] tls connection terminated frequently when mongodb cluster member becomes unavailable (#1599)

by flo-eyeson-team

### Description Kamailio frequently terminates tls connections to sip clients, if db_mongodb database driver is used and one mongodb cluster member becomes unavailable. If one mongodb secondary is shutdown (or the primary and a promotion takes place) kamailio will terminate the tls connection to a sip-client (sends a FIN/ACK) after a short period of time (differs 20s to 180s). If a sip client reconnects, registration will succeed (so mongodb access is fine) but kamailio will terminate the tls connection shortly after. This situation happens as well if only the connection between kamailio and a secondary is blocked (the cluster members can still communicate with each other). I assume that the mongo-c's server discovery might be the culprit. Mongoc iteratively tries to connect to all servers in the mongoc-uri to update their status (`mongodb://mongodb-cluster2:27017,mongodb-cluster1:27017,mongodb-cluster0:27017/kamailio?authMechanism=MONGODB-X509&replicaSet=rs1&ssl=true&sslcertificateauthorityfile=mongo.ca&sslclientcertificatekeyfile=mongo.certkey`). Nonethless, mongoc's server discovery should be transparent to the kamailio client as long as a primary exists (and this is the case). No reason to terminate tls client connections.  ### Troubleshooting #### Reproduction  1. Setup a mongodb cluster with 3 members. 2. Setup kamailio with tls using db-mongodb via tls. Note: The mongodb-uri must contain all three members, e.g. `mongodb://mongodb-cluster2:27017,mongodb-cluster1:27017,mongodb-cluster0:27017/kamailio?ssl=true`) 3. Connect/register a sip client via tls. 4. Stop one member. 5. Clients will receive a ready error on its tls connection. #### Log Messages  The db_mongodb kamailio logs don't reveal any errors. Operations are normal. The only thing which is shown in the logs is TLS read:error. As seen below. ``` Jul 23 14:22:09 bda3e8fce481 /usr/local/sbin/kamailio[318]: ERROR: tls [tls_util.h:42]: tls_err_ret(): TLS read:error:140E0197:SSL routines:SSL_shutdown:shutdown while in init Jul 23 14:22:09 bda3e8fce481 /usr/local/sbin/kamailio[318]: ERROR: <core> [core/tcp_read.c:1485]: tcp_read_req(): ERROR: tcp_read_req: error reading - c: 0x7fa9b6343210 r: 0x7fa9b6343290 ``` A wireshark trace reveals that kamailio closes the tcp connection (FIN/ACK). The client receives a "stream truncated". ### Additional Information * **Kamailio Version** - output of `kamailio -v` ``` version: kamailio 5.1.4 (x86_64/linux) flags: STATS: Off, USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC, Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLACKLIST, HAVE_RESOLV_RES ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: unknown compiled on 13:39:15 Jul 23 2018 with gcc 6.3.0 ``` * **mongoc-version** ``` mongoc-1.11.0 (built with cmake ../ -DENABLE_AUTOMATIC_INIT_AND_CLEANUP=OFF) ``` * **Operating System**:  ``` Debian 9.5 docker image running on CentOS 7 (kernel 3.10.0-514.6.1.el7.x86_64) ``` -- 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/issues/1599

1 year

[kamailio/kamailio] core: use standard c atomic code from stdatomic.h instead of providing assembly routines (#1430)

by Victor Seva

>From https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=886111 ``` Source: kamailio Version: 5.1.0-1 Severity: wishlist Hi, I noticed from the build log on mips64el a lot of warnings like this: > In file included from ../../core/parser/../mem/../atomic/atomic_native.h:50:0, > from ../../core/parser/../mem/../futexlock.h:42, > from ../../core/parser/../mem/../lock_ops.h:75, > from ../../core/parser/../mem/shm.h:39, > from ../../core/parser/../mem/shm_mem.h:34, > from ../../core/parser/../ut.h:45, > from ../../core/parser/../ip_addr.h:39, > from ../../core/parser/msg_parser.h:37, > from app_sqlang_api.h:28, > from app_sqlang_kemi_export.c:32: > ../../core/parser/../mem/../atomic/atomic_mips2.h:41:2: warning: #warning mips64 atomic code was not tested, please report problems to serdev(a)iptel.org or andrei(a)iptel.org [-Wcpp] > #warning mips64 atomic code was not tested, please report problems to \ > ^~~~~~~ If possible, kamailio should use the standard c atomic code from stdatomic.h instead of providing assembly routines. This will improve maintainability and architecture support. Thanks, James ``` -- 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/issues/1430

1 year

[kamailio/kamailio] tmx: t_reuse_branch() doesn't copy $du into new branch (#1264)

by Sebastian Damm

### Description When using the t_reuse_branch() function after a branch failure, Kamailio doesn't send out the call again if $du was filled and differed from $ru. Instead, Kamailio tries to send the call directly to $ru. I would expect that the call gets sent out the same way as the original call. ### Troubleshooting #### Reproduction You need a setup like this: UAC1 K(LB) <=> K(Proxy/Registrar) <=> K(LB) => UAC2 When a UAC registers, the Path module is used to save the LB (Loadbalancer) for outbound calls. When UAC1 calls UAC2, the original call gets sent through K(LB) to UAC2. However, if the first call gets rejected and the call gets retried through branch failure route, K(Proxy/Registrar) tries to send the call directly to the request URI. In your branch_route add a reference to a branch failure route: ```t_on_branch_failure("samplename");``` Then add a branch failure route like this: ```event_route[tm:branch-failure:handle_tls_branch] { if(t_check_status("488")) { if (isbflagset(0) && !isbflagset(1)) { t_reuse_branch(); setbflag(1); t_relay(); } } }``` When filling $du manually before sending out the call again, everything works. ### Possible Solutions The original destination URI should be copied to the reused branch. ### Additional Information * **Kamailio Version** - output of `kamailio -v` ``` version: kamailio 4.4.5 (x86_64/linux) flags: STATS: Off, USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC, Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX, FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR, USE_DST_BLACKLIST, HAVE_RESOLV_RES ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB poll method support: poll, epoll_lt, epoll_et, sigio_rt, select. id: unknown compiled with gcc 4.7.2 ``` But from what I see in source code, the devel version still copies the same value. * **Operating System**:  ``` Debian Wheezy ``` -- 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/issues/1264

1 year

[kamailio/kamailio] Solve the high potential of having conflicts for internal dialog id in multiple simultaneous kamailio instances (#1500)

by alexsosic

Regarding the issue 1274 (https://github.com/kamailio/kamailio/issues/1274) there's a problem in highly scalable environments to avoid conflicts in storing dialogs in the same DB. As @miconda pointed out: > The internal id is made from two integers, hash entry and hash id (h_entry, h_id). h_entry is the index of the slot in the hash table used to store the dialog structure, computed by hashing over call id. h_id is an incremented integer specific for each hash table slot. > > So, when loading a new record from database, if its h_id is not conflicting with an existing dialog on the same hash table slot, all is ok, otherwise the module is not going to work properly with two dialogs having same h_id. > > So, when loading a new record from database, if its h_id is not conflicting with an existing dialog on the same hash table slot, all is ok, otherwise the module is not going to work properly with two dialogs having same h_id. > > Among solutions I thought of: > > 1. have the servers generating non conflicting h_id, by having a start value different per server and an increment step larger than the number of servers. Iirc, here was at some point a similar attempt, but somehow didn't make it. Let's say one has two servers, first server starts allocating h_id from 1 and increments by 2 (e.g: 1, 3, 5, ...) and the seconds start from 2 and increments with 2 (2, 4, 6, ...). Those values can be set via mod params, eventually with an option to rely on server_id. > This should be the least intrusive in the other modules built on top of dialog. But it is rather rigid, with the example above, if one adds an extra server, it needs to reconfigure the old ones, so each server starts from either 1, 2 or 3 and increment by 3. Of course, one can set increment step to a larger value, like 100, and then has flexibility to deploy up to 1 hundred servers before having to reconfigure in case there is need for more server. > > 2. add server_id as the third field in the dialog id. It will require review of other modules using dialog and eventual code updates in those modules. One column to store the server_id needs to be added to dialog db tables. > > 3. switch from this conflicting id system to something like string unique ids, similar to what we have in usrloc records. This will require coding in other modules, changes to database schema, etc., so more work comparing with the above two, but could be the best in long term. It would be great to have the option 3 developed if possible but we would also be satisfied with the option 2 to have a third field, server_id, in the dialog id. Thank You! -- Alex -- 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/issues/1500

1 year

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

sr-dev October 2019