-
Notifications
You must be signed in to change notification settings - Fork 43
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fax is not hanging up after successful completion #31
Comments
As an additional info. This happens only when T38 support is switched on my router/sip gateway. If T38 is switched off the fax hangs up correctly after a successfully sent fax. If T38 support is switched on i can see that leg A (internal connection from endpoint behind router to router/sip gateway) is transmitted with T38 but on leg B (external connection to fax somewhere in the world) the connection is running without T38. In that scenario the fax is also sent successfully but the connection remains open forever and fax/freeswitch is not hanging up. |
Hello. Great project! mod_spandsp_fax.c:276 FLOW T.30 Success - delivered 1 pages Yet, faxstat -s shows: HylaFAX scheduler on mwr: Running JID Pri S Owner Number Pages Dials TTS Status Here, restarting Hylafax resends the fax as it thinks it was interrupted. A work-around is to disable T.38 in /etc/gofax.conf and to restart the stack. I tried this on another server with another ISP and used a non-standard port (5080 instead of 5060) for SIP communication and it worked sporadically. I think it is a bug in freeswitch or in gofaxip. Unfortunately my skills are not enough to try to provide a patch. Thanks again for an excellent software product, running well with T.30. |
I encountered similar problems with the latest Freeswitch Stable (1.6.19) in combination with a Patton Smartnode... After transferring the fax, the Smartnode tries to re-invite to g.711 again. With Freeswitch 1.6.19, the re-invite to g.711 is answered with 200 OK, and then the dialog hangs... I downgraded to 1.6.15 and deactivated the fax detection after 1 second on the patton (could have been sufficent alone!), that worked for me. We will have a look on that and ensure, gofax.ip will work properly with the current version of Freeswitch. |
I think the reason for this are: |
Is there already a solution resp. the real reason known? I've a setup where I can reproduce this behavior by sending a fax (via our sip provider) to myself.
If you need detailed logs or other information feel free to ask for it. |
Did anyone found a solution for that, it happens to me once a day, it's usually on jobs with more than 10 pages.... |
Sorry, but that never happens to us and we send 20+ pages faxes on average,
more than 1000 faxes daily without any issue.
…On Thu, Nov 2, 2017 at 1:38 PM, chaim1989 ***@***.***> wrote:
Did anyone found a solution for that, it happens to me once a day, it's
usually on jobs with more than 10 pages....
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#31 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/ABK1huO-4Yth2G3WR9k5RSlLt9SVjdQ7ks5sycXCgaJpZM4NFZUX>
.
|
can you give me the details of your enviroment, |
We have 2 FS with gofaxip, each with 60 'modems', both are VMWare instances
with 8Gb of RAM and 2 VCPU @ 3.2Ghz
They send/receive the FAX thought a SipWise NGCP 4.5.2 platform, not
virtualized, our carriers used mostly Sonus equipments.
…On Fri, Nov 3, 2017 at 1:57 AM, chaim1989 ***@***.***> wrote:
can you give me the details of your enviroment,
i will be happy to find an enviroments where its working perfect,
os, hardware/virtualization, etc.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#31 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/ABK1hld8EnCD5mA5Jns9eRVdtZ26LO7Oks5synMOgaJpZM4NFZUX>
.
|
@chaim1989 you can try to build and run the version of PR #37. probably that solves the issue for you (and if it makes it worse feel free to report that in the PR) |
this is the input of my faxstat -s command: 2340 127 R fax 97248600303 14:14 1:4 Sending 14400 it's just get stuck like that at least twice a day, usually on jobs that are more then 5 pages long, it only releases when i do "systemctl restart freeswitch" any ideas ? does it related to this issue of fax not hangingup |
okay, then you're having a different issue than me, sorry :) |
I don't mind changing hardware or using different Linux version I just want
one that works 100%.
…On Nov 13, 2017 17:06, "Steffen Jaeckel" ***@***.***> wrote:
i did change to the PR version of sjaeckel (i just replaced the binary of
gofaxd)
okay, then you're having a different issue than me, sorry :)
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#31 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AKQ4O5cyY-iZDu7OjWH5BcX4SbjI8dRZks5s2FsAgaJpZM4NFZUX>
.
|
I did encounter some changes in freeswitchs behaviour since 1.6.15, so you could try to use 1.6.13 (that worked for me a long time) to see if that changes anything... Anyway... please provide a fs_cli trace with:
|
I am able to reproduce your problem and i will open an issue in freeswitch (propably tomorrow) because from my point of view freeswitch should terminate the call after calling txfax when txfax is the only application triggered by originate... But i can only reproduce the problem when the other side of the fax transmission is another gofaxip server without the missing hangup from #37 ... Do you know anything about the receiving side in your scenario? Because normally i would expect the other side to be interested in terminating the call after the transmission too ;) |
It happens in different receiving point nothing common between them,
But it is definitely a free switch issue,
to release the stuck jobs I just restart freeswitch
…On Dec 4, 2017 7:47 PM, "Sebastian Denz" ***@***.***> wrote:
I am able to reproduce your problem and i will open an issue in freeswitch
(propably tomorrow) because from my point of view freeswitch should
terminate the call after calling txfax when txfax is the only application
triggered by originate...
But i can only reproduce the problem when the other side of the fax
transmission is another gofaxip server without the missing hangup from #37
<#37> ...
Do you know anything about the receiving side in your scenario?
Because normally i would expect the other side to be interested in
terminating the call after the transmission too ;)
If it actually is another gofaxip server, your problem should be gone if
the fix from #37 <#37> is deployed
on the receiving end...
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#31 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AKQ4O5I1JD_MG0UqDPKzGLNDu7zHs12jks5s9DAagaJpZM4NFZUX>
.
|
I just opened an issue in freeswitch: If you have any further information feel free to add them to the issue! |
Denzs thanks for taking care of it, |
As mentioned before, i am using 1.6.15 without problems.... At this very moment i am building 1.6.16 and 1.6.17 to narrow it down... In most cases i am using gofaxip with Patton Smartnode media-gateways, these devices do send a re-invite back to g711 after the fax transmissions which results in freeswitch (1.6.15) sending a -> this leads to the unsatisfying hangup cause INCOMPATIBLE_DESTINATION but at least it works! I am not sure, how freeswitch 1.6.15 behaves when the other side doesnt send a re-invite... I suggest you try 1.6.15, and if that doesnt help please provide a sofia siptrace log, then we have to check if we can do anything about it... |
Thanks a lot i will let you know
<https://mailtrack.io/> Sent with Mailtrack
<https://chrome.google.com/webstore/detail/mailtrack-for-gmail-inbox/ndnaehgpjlnokgebbaldlmgkapkpjkkb?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality>
…On Wed, Dec 6, 2017 at 1:34 PM, Sebastian Denz ***@***.***> wrote:
As mentioned before, i am using 1.6.15 without problems....
At this very moment i am building 1.6.16 and 1.6.17 to narrow it down...
In most cases i am using gofaxip with Patton Smartnode media-gateways,
these devices do send a re-invite back to g711 after the fax transmissions
which results in freeswitch (1.6.15) sending a
SIP/2.0 488 Not Acceptable Here followed by a hangup initiated by
freeswitch.
-> this leads to the unsatisfying hangup cause INCOMPATIBLE_DESTINATION
but at least it works!
I am not sure, how freeswitch 1.6.15 behaves when the other side doesnt
send a re-invite...
I going to test that later...
I suggest you try 1.6.15, and if that doesnt help please provide a sofia
siptrace log, then we have to check if we can do anything about it...
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<https://mailtrack.io/trace/link/3216f8a19c08c87a2ce4132662be4e42ad670eda?url=https%3A%2F%2Fgithub.com%2Fgonicus%2Fgofaxip%2Fissues%2F31%23issuecomment-349613286&userId=2203333&signature=c625329c5b4fc607>,
or mute the thread
<https://mailtrack.io/trace/link/fddc47c2ac385ed5faed5aad8df9d524c31ea8d5?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAKQ4O7wj58-EGYiawOzg3zUdr6jmVtlnks5s9nvIgaJpZM4NFZUX&userId=2203333&signature=9bd23f25de561ab5>
.
|
After testing:
|
I can confirm that fs 1.6.16 works as expected, |
Do you have a comparable setup with Re-Invite to G711 after the transmission followed by 488 and hangup or do you have a "clean" call shutdown? |
I am on 1.10.6 on Raspberrypi / Buster and unfortunately hanging up does not work as expected. |
Wouldn't it be possible to fix the code of gofax.ip so that it also works with current versions of freeswitch? Staying on a legacy version forever doesn't sound like a good idea, I have to say. |
I am on 1.10.7 and getting the same issue of freeswitch not hanging up after the fax transmission completes. I have an old system on 1.6 where it is working fine, on the same network (different RTP ports), so looks like an new issue. |
@JonathanAldridge I have the same issue, with the same version. Can you please help me on this? |
Can you provide us some logs @jigsp26 ? |
4b58dd51-f6eb-4287-bde1-71434a952524 2022-07-29 13:41:04.779861 91.70% [DEBUG] switch_core_state_machine.c:323 sofia/external/8175495330 Standard EXECUTE After successfully send, calls will be send after 3 min |
I’ve had no luck with this, and have continued running the 1.6 version.
JA
|
I have the same problem. What logs do you need? |
The reason for this issue is a change in behaviour of FreeSWITCH/mod_spandsp.. As i mentioned in #44 - the problems were introduced with FreeSWITCH 1.6.17. I already opened a bug (> 4 years ago) in the old issue tracker, but as the issues were moved from Atlassian Tools to Github, the issue was closed... But there already seem to be a "new version" of the bug: signalwire/freeswitch#1640 This definitely has to be fixed in FreeSWITCH/mod_spandsp.. i dont see any way to workaround this in an acceptable way :-/ |
Still the issue is open, I can't use any other version |
As i already said.. this has to be fixed upstream inside of FreeSWITCH... :-/ |
Was a solution ever found for this issue with FreeSwtich? V 1.10.7 still seems to have it in some call cases. |
Unfortunately, my last comment is still up to date... :-/ |
SO downgrading to 1.6.15 works correctly with t38? |
1.6 should work right.. i know thats an unsatisfying answer :-/ |
1.6.? Thanks for the insight.
…On Thu, Jan 19, 2023, 5:22 AM Sebastian Denz ***@***.***> wrote:
1.6 should work right.. i know thats an unsatisfying answer :-/
—
Reply to this email directly, view it on GitHub
<#31 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAIFHFT4P4WI44CFRCNHVU3WTEWX5ANCNFSM4DIVSULQ>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
1.6.16 should be good |
Thanks!
…On Thu, Jan 19, 2023, 9:02 AM Sebastian Denz ***@***.***> wrote:
#31 (comment)
<#31 (comment)>
1.6.16 should be good
—
Reply to this email directly, view it on GitHub
<#31 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAIFHFXI3CZ4FGXMNTUI6CTWTFQSRANCNFSM4DIVSULQ>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
Doesnt seem this can be compiled with AWS supported flavors of linux anymore. Any additional ideas? |
@Brettk80 please open another issue for this and do not forget to add some information about the concrete problem :) |
1.6.x was complicated. We still see long calls where FS doesnt hangup on FreeSWITCH Version 1.10.11-release~64bit (-release 64bit) I'd have to do more digging but my suspicion is these stuck calls are voice and not fax and reach the talk limit without a valid carrier in some cases. It appears a bye is received but maybe not heard or honored. |
Hi Team,
i have installed gofaxip and freeswitch as per your instructions and started testing it. After some tries and adjustments in the config i could create successfully my first faxes with Hylafax and freeswitch, gofaxip however in almost all the cases the fax keeps the voip connection open forever even if spandsp reports through the event socket that the fax was successfully delivered. I have to shutdown freeswitch to get the fax to hang up and complete. This happens in almost all my test cases. See one log attached:
2017/04/22 23:23:07.052553 Processing HylaFAX Job 440 as e1062558-b04b-4fc9-8a58-5336b6511b2b
2017/04/22 23:23:07.089390 Originating channel to (FAX NUMBER) using gateway default
2017/04/22 23:23:13.932231 Originate successful
2017/04/22 23:23:13.932649 Call state change: EARLY
2017/04/22 23:23:13.933085 Call state change: ACTIVE
2017/04/22 23:23:21.784950 Remote ID: "", Transfer Rate: 14400, ECM=true
2017/04/22 23:23:34.373968 Page 1 sent: Image Size: 1728x1162, Compression: T.6, Comp Size: 1701 bytes, Bad Rows: 0
2017/04/22 23:23:54.439536 Page 2 sent: Image Size: 1728x1162, Compression: T.6, Comp Size: 26519 bytes, Bad Rows: 0
Hangs up only because i shutdown freeswitch manually after waiting another 6 mins to see if fax will hang up automatically.
2017/04/22 23:29:27.862314 Call state change: HANGUP
2017/04/22 23:29:27.865432 Success: true, Hangup Cause: SYSTEM_SHUTDOWN, Result: OK
Regards
Rob
The text was updated successfully, but these errors were encountered: