### Description I have a problem with publish/notify messages generated from Kamailio towards a monitoring device after multiple branches has been created. When the call is forked by kamailio, invites are generated towards each callee user agent depending on how many branches we have in the call. When one of the callee user agents sends back 200 ok to enter the conversation, kamailio automatically sends a CANCEL towards the user agents that lost the race condition and didnt take the call. When this happens, a notify is set to the monitoring user agent with the state "terminated" but this is wrong since the call is up and we are in conversation.

#### Reproduction

1. set up a monitoring device that subscribes to B events. 2. Register more than one device to B to make kamailio behave like a forking proxy 3. Call B and observe the notify message sent to the monitoring device, in particular the <dialog id="xxx" value

### Possible Solutions

I think this can be solved just adding some random uuid inside the xml field sent inside the xml of the NOTIFY message: The random uuid must be different for each branch.

### Additional Information

version: kamailio 5.5.1 (x86_64/linux) flags: USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS, USE_RAW_SOCKS, DISABLE_NAGLE, USE_MCAST, NO_SIG_DEBUG, DNS_IP_HACK, 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_BLOCKLIST, HAVE_RESOLV_RES, TLS_PTHREAD_MUTEX_SHARED 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 10.2.1

* **Operating System**:

No LSB modules are available. Distributor ID: Debian Description: Debian GNU/Linux 11 (bullseye) Release: 11 Codename: bullseye

Linux sp1 5.10.0-9-amd64 #1 SMP Debian 5.10.70-1 (2021-09-30) x86_64 GNU/Linux