Best Of
SIP REFER and SIPREC
this is on an ESBC 9.3
call comes in from our PSTN carrier and is sent via local policy to a system on a TLS/SRTP trunk. This call does not need to be (should not be) recorded.
Now, some of the calls are transferred back, via SIP REFER, and then go to a different system and this leg of the call needs to be recorded (SIPREC).
I can successfully route the call to its new destination (if it matters, I have refer-src-routing disabled) but I cannot get it to start the SIPREC. I'm thinking this is because the original call, before it came back via the REFER, never was recorded - ?
is there a way to enable SIPREC for the new leg of the call, after the SIP REFER?
thank you
SBC HA when using SIP/TCP
Assuming the standard deployment of a call server setting up calls between sip phones, the SBC separating the two:
- sip phones on one side of the SBC - access
- call server on the other side of the SBC - core.
SIP/UDP is used between the SBC and the call server.
As the sip phones are behind NAT, HNT is used.
When SIP/UDP is used between the phones and the SBC, in case of a SBC HA failover, the active calls are maintained and if one side disconnects the call, the call is properly disconnected on the remote side too.
However, if I use SIP/TCP between the sip phones and the SBC, in case of a SBC HA failover, the calls are maintained but if one side disconnects the call, the remote side is not disconnected.
Traces shows that the call server receives the BYE message from the disconnecting side and send a BYE message to the remote side - via the SBC. However, the SBC replays 503 service unavailable - looks like if it doesn't know how to forward the call to the remote side - as if he lost the TCP connection.
Is this the expected behaviour from the SBC HA when using SIP/TCP ?
If not, any idea how I can get calls using SIP/TCP to behave like the calls using SIP/UDP and get disconnected properly ?
Tnx,
Dan
Re: Since RU 19.28 cdump twice a day without incident
Patch is ready to download for Linux platform
Re: ODA backup report and arhivelog auto backup
problem solved… I do not know why… but at least something worked…
Re: Maintenance schedule PUM content
Hi,
Documentation for PUM Image 53 once it will be available, will pe posted in the below document:
PeopleSoft Update Manager (PUM) Home Page (Doc ID 1641843.2) > PeopleSoft Update Image Home Pages
Regards,
Re: How does Oracle AP6350 SBC select local-policy used to route a call?
Hello
as already told u in the previous post LP-A is matching because To-address has priority over From-address , and in this case To-address is matching that ip address .
For this type of scenario what u can do is to configure cost=1 on LP-A , leaving cost=0 on LP-B , but I'm unaware of other type of traffic that u need to handle .
Cheers
Antonio
Re: ODA X11 new deployment
Hi Tamil,
Thanks for the feedback.
We have ascertained that 23ai is not supported on BM but only on DBsystems/KVM.
A cleanup has been done and re-create appliance but failed because of customer's NTP is not working.
Once the NTP is resolved, the appliance will be re-created.
Thanks for your support.
Thanks,
James.
Re: Enter Feedback Here for the EBS Support Analyzer Bundle Menu Tool - Note 1939637.1
Hi Surya,
I discussed your error with our DBA expert and he offered
Oracle 8i SQL*Plus limits
While the definitive maximum statement length in Oracle is limited by several factors including database configuration, available disk space, and memory, this article states that there is no fixed number. However, it is possible to list several relevant limits related to SQL*Plus in Oracle 8i:
Additionally, after each error in your log shows :
Package body created.
No errors.
INFO: Dependency File exited with status: 0
so I think you are safe to ignore these.
To be sure, confirm that you are able to run the TITLE: Inventory Counting Analyzer (one example) from the bundle and as a concurrent request.
Let me know if you see a different message if you encounter this error.
Cheers,
Bill.
Re: oracle patch downlaod error
I have the same, and my temporary fix is still : use another type of browser. You should always have at least 2 types of browsers (and one of them not Microsoft).
Mr. C