Hi Earl,

Nasty, I recognize that bug.  The configuration data is being
pre-interpreted before my scripts see it so the URLs are all wrong.
I've sent a patch off to Epiphan, but for now I've built a patched
version which is available at
http://aries.usask.ca/epiphan/frm-2-3-0e-Matterhorn-15977_1067-VGA2CPU_DM365.bfrm.

G


On 12-03-26 02:15 PM, Earl Schellhous wrote:
> Hi Greg,
> 
> Unfortunately we are behind a firewall but if you are willing to work with me 
> remotely, I can get us a "room" in Adobe Connect and I can turn over remote 
> desktop access to you. If you want to do that, we should probably take the 
> conversation offline from the matterhorn-users list. My email is 
> [email protected] and my phone is (530) 754-9346. If you have a mic 
> headset we can use VoIP to talk in Adobe Connect.
> 
> At the calendar URL /recordings/calendars?agentid=&cutoff=1334001970000 there 
> is some text output:
> BEGIN:VCALENDAR
> PRODID:Opencast Matterhorn Calendar File 0.5
> VERSION:2.0
> CALSCALE:GREGORIAN
> END:VCALENDAR
> 
> at this url /capture-admin/agents.json the time-since-last-update field never 
> goes down in value, on every refresh it goes to a higher number.
> 
> Here is the ECD log during an attempt that failed:
> Mar 26 12:56:21 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 12:57:23 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 12:58:25 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 12:58:29 MTH92151 user.info MH.sendCaps: Capabilities sent to 
> /capture-admin/agents//configuration
> Mar 26 12:59:30 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 13:00:32 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 13:01:34 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 13:01:56 MTH92151 user.debug MH.Scheduler[911]: Fetching calendar from 
> /recordings/calendars?agentid=&cutoff=1334001716000
> Mar 26 13:01:56 MTH92151 user.notice MH.Scheduler[911]: Calendar not updated, 
> using cached at /data/u/matterhorn/cached_schedule.ics instead
> Mar 26 13:01:58 MTH92151 user.debug MH.Scheduler[911]: Unable to schedule 
> event 2865 in the past.
> Mar 26 13:01:58 MTH92151 user.debug MH.Scheduler[911]: No change in the 
> crontab, skipping write.
> Mar 26 13:02:36 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 13:02:39 MTH92151 user.info MH.sendCaps: Capabilities sent to 
> /capture-admin/agents//configuration
> Mar 26 13:03:00 MTH92151 user.debug MH.Scheduler[911]: Fetching calendar from 
> /recordings/calendars?agentid=&cutoff=1334001780000
> Mar 26 13:03:00 MTH92151 user.notice MH.Scheduler[911]: Calendar not updated, 
> using cached at /data/u/matterhorn/cached_schedule.ics instead
> Mar 26 13:03:02 MTH92151 user.debug MH.Scheduler[911]: Unable to schedule 
> event 2865 in the past.
> Mar 26 13:03:02 MTH92151 user.debug MH.Scheduler[911]: No change in the 
> crontab, skipping write.
> Mar 26 13:03:41 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 13:04:04 MTH92151 user.debug MH.Scheduler[911]: Fetching calendar from 
> /recordings/calendars?agentid=&cutoff=1334001843000
> Mar 26 13:04:04 MTH92151 user.notice MH.Scheduler[911]: Calendar not updated, 
> using cached at /data/u/matterhorn/cached_schedule.ics instead
> Mar 26 13:04:05 MTH92151 user.debug MH.Scheduler[911]: Unable to schedule 
> event 2865 in the past.
> Mar 26 13:04:05 MTH92151 user.debug MH.Scheduler[911]: No change in the 
> crontab, skipping write.
> Mar 26 13:04:43 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> Mar 26 13:05:07 MTH92151 user.debug MH.Scheduler[911]: Fetching calendar from 
> /recordings/calendars?agentid=&cutoff=1334001907000
> Mar 26 13:05:07 MTH92151 user.notice MH.Scheduler[911]: Calendar not updated, 
> using cached at /data/u/matterhorn/cached_schedule.ics instead
> Mar 26 13:05:09 MTH92151 user.debug MH.Scheduler[911]: Unable to schedule 
> event 2865 in the past.
> Mar 26 13:05:09 MTH92151 user.debug MH.Scheduler[911]: No change in the 
> crontab, skipping write.
> Mar 26 13:05:45 MTH92151 user.info MH.sendState: Agent set to idle at 
> /capture-admin/agents/
> 
> 
> Earl
> 
> 
> Earl Schellhous
> Programmer
> IET Academic Technology Services
> (530) 754-9346
> 
> 
> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]] On Behalf Of Greg Logan
> Sent: Monday, March 26, 2012 12:35 PM
> To: [email protected]
> Subject: Re: [Matterhorn-users] Epiphan capture appliance stuck in 'Capturing'
> 
> On 12-03-26 01:29 PM, Earl Schellhous wrote:
>> Hi Greg,
>>
>> I'm working with Charlie on this issue and we have done the following:
>> 1) Installed 2.3.0e firmware
>> 2) restored to factory settings
>> 3) reset the critical settings (matterhorn server etc.)
>> 4) changed the ingest interval to 5 minutes
>> 5) attempted 1 minute capture.
>>
>> The result was the same as before, failure giving this message:
>> "WARNING: Recording may have failed to start or ingest! It seems the core 
>> system did not receive proper status updates from the Capture Agent that 
>> should have conducted this recording."
>>
>> What can we try next?
> 
> Double check that the settings on the ECD are correct: core url includes a 
> port, username and passwords are correct.  What does the log on the device 
> say?  It should include the exact URL from which it's trying to retrieve the 
> calendar data.  Does that URL work from another machine?
> When you go to $Your_Core_url/capture-admin/agents.json (eg:
> http://testadmin.usask.ca:8080/capture-admin/agents.json) what does the 
> "time-since-last-update" field say for that agent?  It should reset down to 0 
> every time the agent sends a state update.
> 
> If those are all right then can I get at the device remotely?  I can take a 
> look internally and see what, if anything, is attempting to capture.
> 
> G
> 
>> Thank you!
>> Earl
>>
>> Earl Schellhous
>> Programmer
>> IET Academic Technology Services
>> (530) 754-9346
>>
>> -----Original Message-----
>> From: [email protected] 
>> [mailto:[email protected]] On Behalf Of 
>> Greg Logan
>> Sent: Sunday, March 25, 2012 11:33 AM
>> To: [email protected]
>> Subject: Re: [Matterhorn-users] Epiphan capture appliance stuck in 
>> 'Capturing'
>>
>> On 3/23/2012 12:34 PM, Charles Turner wrote:
>>> Hello,
>>>  
>>> We're new to this list and are experiencing problems with Epiphan's 
>>> Matterhorn capture appliance.  From earlier list traffic it seems 
>>> like others have experienced similar problems, but it isn't clear to 
>>> us what the current state is. We are using Epiphan firmware 2.2.3 and 
>>> Matterhorn server 1.3.0 (details below).
>>>  
>>> The problem is that we schedule a recording (1 minute) and when the 
>>> recording is finished it remains in "Capturing" mode indefinitely.
>>>  Clicking on the Matterhorn Admin tools -> Inspect Workflow screen 
>>> shows that it is in a PAUSED state.
>>>  
>>> If we then go into the /admin/maintenancecfg (maintenance page) of 
>>> the Epiphan device admin and click on "Reboot Now" after a few 
>>> minutes the recording changes out of the paused state and begins 
>>> processing normally and when finished we can click on "Play" and the video 
>>> will play.
>>
>> This, to me, says that you aren't waiting long enough after it stops 
>> capturing.  The ingest script runs every 60 minutes by default, so if you 
>> capture a one minute lecture right after the ingest script runs then you 
>> have to wait another 59 minutes before the ingest script fires again.  The 
>> script also fires on reboot, so I'm betting that's why you see it suddenly 
>> start working after restarting the box.
>>
>>> This is true of Epiphan firmware 2.2.1a and 2.2.3. 
>>>
>>> If we use firmware 2.3.0e we get the following message instead of
>>> "Capturing":
>>> "WARNING: Recording may have failed to start or ingest! It seems the 
>>> core system did not receive proper status updates from the Capture 
>>> Agent that should have conducted this recording."
>>
>> Two options:  Either the capture did not start properly or the device did 
>> not update its calendar before the capture should have started.
>> Again, like the ingest script, the calendar script fires every X minutes.  
>> If you schedule something to start sooner than X minutes out then the device 
>> won't start capturing since the event will have started in the past 
>> (mid-event starts are still coming).
>>
>> For what it's worth, I've been working with 2.3.0e (which should be a full 
>> release, just hasn't been copied to the right place) all week verifying some 
>> memory issues and things have been working fine for me.
>> Have you been resetting the device to factory defaults when you change 
>> firmware versions?
>>
>> G
>>
>>> Does anybody know of a way to resolve the issue, or are we still 
>>> waiting on a new release of the firmware from Epiphan?
>>>
>>> Configuration details:
>>>
>>> Matterhorn svn info:
>>> Last Changed Author: greg_logan
>>> Last Changed Rev: 12007
>>> Last Changed Date: 2012-03-14 21:52:05 -0700 (Wed, 14 Mar 2012)
>>>  
>>> From the admin/info page of the device:
>>> Matterhorn MTH92145 info
>>> FIRMWARE_VERSION="2.2.3"
>>> FIRMWARE_BUILD_HOST="tochilka 2.6.38-8-server"
>>> FIRMWARE_BUILD_DATE="2011-11-27"
>>> FIRMWARE_PLATFORM=VGA2CPU_DM365
>>> FIRMWARE_ARCH=arm
>>> FIRMWARE_SVNPATH=/branches/VGABROADCASTER_DAVINCI/2_2_3/epiphan/vga2w
>>> e
>>> b/subsys/base
>>>  
>>> We have tried firmware 2.2.1a, 2.2.3 and 2.3.0e and are currently 
>>> using 2.2.3.
>>>
>>> Thank you!
>>> Charlie Turner
>>> Earl Schellhous
>>> IET-Academic Technology Services
>>> UC Davis
>>>
>>>  
>>>
>>>
>>> _______________________________________________
>>> Matterhorn-users mailing list
>>> [email protected]
>>> http://lists.opencastproject.org/mailman/listinfo/matterhorn-users
>>
>> _______________________________________________
>> Matterhorn-users mailing list
>> [email protected]
>> http://lists.opencastproject.org/mailman/listinfo/matterhorn-users
>> _______________________________________________
>> Matterhorn-users mailing list
>> [email protected]
>> http://lists.opencastproject.org/mailman/listinfo/matterhorn-users
> 
> 
> _______________________________________________
> Matterhorn-users mailing list
> [email protected]
> http://lists.opencastproject.org/mailman/listinfo/matterhorn-users


Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
Matterhorn-users mailing list
[email protected]
http://lists.opencastproject.org/mailman/listinfo/matterhorn-users

Reply via email to