I got a Pro and re-set everything up. From my experience with the first install and then this one, nothing is going to work right when Channels is first pulling the guide, to include any recordings you have scheduled. After the initial pull I haven’t seen any issues (on my OG and on my Pro. I set up Plex with my HDHomeRun’s the other day and it did the same thing (but for a much longer time) so I don’t think it’s a Channels thing, it could be an Android thing.
1 Like
Latest on 16gb Shield:
Haven’t had a drop since reformat/reinstall 
Thanks to @karlmitchell for suggestion and @tmm1 for the nudge.
-
1 Recording this am, no problem. Comskip OK
-
3 back-to-back recordings on 6 different channels
Recordings OK
Comskip failed on recordings 3,4,5. (1,2 &6 are OK)
Note: nothing else running on Shield during this time.
Extract of log:
2017/05/30 14:00:02 [DVR] Running commercial detection on file 183 (TV/Secrets/2017-05-30-1300 Secrets 2017-05-29 S04E06 Shrunken Heads.mpg)
2017/05/30 14:18:23 [DVR] Commercial detection finished with 10 markers.
2017/05/30 14:18:24 [DVR] Processing file-182: TV/Days of our Lives/2017-05-30-1300 Days of our Lives 2017-05-30 S52E177.mpg
2017/05/30 14:18:26 [DVR] Running commercial detection on file 182 (TV/Days of our Lives/2017-05-30-1300 Days of our Lives 2017-05-30 S52E177.mpg)
2017/05/30 14:39:02 [DVR] Commercial detection finished with 16 markers.
2017/05/30 14:39:04 [DVR] Processing file-181: TV/Wicked Tuna/2017-05-30-1300 Wicked Tuna 2013-03-31 S02E10 Captain Carnage.mpg
2017/05/30 14:39:05 [DVR] Running commercial detection on file 181 (TV/Wicked Tuna/2017-05-30-1300 Wicked Tuna 2013-03-31 S02E10 Captain Carnage.mpg)
2017/05/30 14:39:33 [DVR] Commercial detection failed with exit status 7
2017/05/30 15:00:00 [TNR] Closed connection to 1318CAEB for ch423
2017/05/30 15:00:00 [DVR] Finished job 1496167200-ch423 CNN Newsroom With Brooke Baldwin
2017/05/30 15:00:00 [DVR] Deleting expired job 1496167200-ch431 General Hospital
2017/05/30 15:00:00 [DVR] Deleting expired job 1496167200-ch434 The People’s Court
2017/05/30 15:00:00 [TNR] Closed connection to 1318CAEB for ch434
2017/05/30 15:00:00 [TNR] Closed connection to 1318CAEB for ch431
2017/05/30 15:00:00 [DVR] Finished job 1496167200-ch434 The People’s Court
2017/05/30 15:00:00 [DVR] Finished job 1496167200-ch431 General Hospital
2017/05/30 15:00:00 [DVR] Processing file-186: TV/The People’s Court/2017-05-30-1400 The People’s Court 2016-11-10 S20E44.mpg
2017/05/30 15:00:02 [DVR] Running commercial detection on file 186 (TV/The People’s Court/2017-05-30-1400 The People’s Court 2016-11-10 S20E44.mpg)
2017/05/30 15:00:21 [DVR] Commercial detection failed with exit status 7
2017/05/30 15:00:21 [DVR] Processing file-185: TV/General Hospital/2017-05-30-1400 General Hospital 2017-05-30 S55E41.mpg
2017/05/30 15:00:22 [DVR] Running commercial detection on file 185 (TV/General Hospital/2017-05-30-1400 General Hospital 2017-05-30 S55E41.mpg)
2017/05/30 15:02:31 [DVR] Commercial detection failed with exit status 7
2017/05/30 15:02:31 [DVR] Processing file-184: TV/CNN Newsroom With Brooke Baldwin/2017-05-30-1400 CNN Newsroom With Brooke Baldwin 2014-06-16.mpg
2017/05/30 15:02:33 [DVR] Running commercial detection on file 184 (TV/CNN Newsroom With Brooke Baldwin/2017-05-30-1400 CNN Newsroom With Brooke Baldwin 2014-06-16.mpg)
2017/05/30 15:10:13 [HLS] Starting transcoder for file-179 at 0s (encoder=h264_mediacodecndk, resolution=1080, deinterlacer=blend, bitrate=10000)
[mpegts @ 0x204185f600] Dropped corrupted packet (stream = 1)
[mpegts @ 0x204185f600] Dropped corrupted packet (stream = 2)
2017/05/30 15:10:51 [HLS] Stopping transcoder session file-179 @ 1m15.408678s
2017/05/30 15:18:43 [DVR] Commercial detection finished with 2 markers.
Tried 3 more. All recorded and completed comskip OK.
However, see log…what is “Buffer” warning. Have never seen before until Shield setup. As internal storage space is limited on 16gb, could this also be a factor in comskip failures?
2017/05/30 17:30:00 [DVR] Starting job 1496179800-ch421 Friends on ch=[421]
2017/05/30 17:30:00 [TNR] Opened connection to 1318CAEB for ch421
2017/05/30 17:30:00 [DVR] Starting job 1496179800-ch425 Say Yes to the Dress on ch=[425]
2017/05/30 17:30:00 [TNR] Opened connection to 1318CAEB for ch425
2017/05/30 17:30:00 [DVR] Starting job 1496179800-ch431 Two and a Half Men on ch=[431]
2017/05/30 17:30:01 [TNR] Opened connection to 1318CAEB for ch431
2017/05/30 17:30:11 [WRN] Buffer for 1318CAEB ch421 is more than 50% full (clients=1, len=16777684)
2017/05/30 17:30:12 [WRN] Buffer for 1318CAEB ch431 is more than 50% full (clients=1, len=16777684)
2017/05/30 17:30:12 [WRN] Buffer for 1318CAEB ch425 is more than 50% full (clients=1, len=16777684)
2017/05/30 17:30:16 [WRN] Buffer for 1318CAEB ch431 is more than 75% full (clients=1, len=25165868)
2017/05/30 17:30:17 [WRN] Buffer for 1318CAEB ch421 is more than 75% full (clients=1, len=25165868)
2017/05/30 17:30:17 [WRN] Buffer for 1318CAEB ch425 is more than 75% full (clients=1, len=25165868)
2017/05/30 17:30:18 [DVR] Recording for job 1496179800-ch421 from 1318CAEB ch421 into “TV/Friends/2017-05-30-1730 Friends 1996-01-28 S02E13 The One After the Super Bowl.mpg” for 29m59.965613884s
2017/05/30 17:30:18 [DVR] Recording for job 1496179800-ch431 from 1318CAEB ch431 into “TV/Two and a Half Men/2017-05-30-1730 Two and a Half Men 2004-02-16 S01E16 That Was Saliva Alan.mpg” for 29m59.03380076s
2017/05/30 17:30:18 [DVR] Recording for job 1496179800-ch425 from 1318CAEB ch425 into “TV/Say Yes to the Dress/2017-05-30-1730 Say Yes to the Dress 2013-09-06 S10E06 Father of the Bride.mpg” for 29m59.332732426s
2017/05/30 17:30:20 [WRN] Buffer for 1318CAEB ch431 is more than 95% full (clients=1, len=31877468)
Note: 3 b-back earlier … all 6 were 1 hour shows
On latest 3 … only 1/2 hour.
Currently recording 3 one hour, then will retry 3 back to back (6 @ 1 hour each) again to try to replicate earlier results.
Buffer warning means disk is not able to write data quickly enough. Could be due to missing power supply, or perhaps too slow disk (USB2?)
Changed out the portable WDPP yesterday before reformat & have since been using WD easystore (supposedly a WD red) 8TB USB3 directly attached to Shield.
Hmm, perhaps some other bottleneck then… does it only occur when 3 simultaneous recordings are happening?
If buffer exceeds 100%, recordings will be corrupted or have missing data.
Given your response and @timstephens24 latest positive results with the Pro, I’m thinking that the 16gb (~8gb free) Shield may not be up to intensive tasks.
Also, if the current process is utilizing the “adopted storage” SDXC, then this card may be causing the problem, if too slow.
After completion of last 3 b-b show test (6 @ 1 hour - on 1st 3 now), I’ll remove it and do a few more tests. Also, I will then attach a smaller powered USB3 as “adopted storage” and test with that. Recordings will still be directed to “removable storage”, powered USB3 8TB, in both cases.
Latest results (seem to support hypothesis):
No problems with 3 half hour shows. Recordings and comskip success all 3.
On 3 full hour shows. Recordings all OK, however 1st file processed failed comskip, the following 2 were successful.
AFAIK the internal storage should not affect the recording process. Data is received from HDHR and written to external storage.
Well that hypothesis was a bust…BUT
Common to ALL comskip failures (on my tests)
720P broadcast
Note: Found that some stations are advertising 1080i, however when analyzing the actual file created…every single one was 720P
My external storage keeps getting unmounted somehow every 3 days (or so), so my log is full of ‘transport endpoint is not connected’ instead of recording what it was supposed to today. Since I’m using a Pro, can you allow saving to the internal storage so I don’t have to deal with this?
I can look into how this would work…
My understanding is that you’d have to store the recordings into the storage area for the app, which also means that all the recordings would be deleted when the app was uninstalled, and it would be quite hard to move the recordings to another system.
All tests done yesteday/today AFTER reset
11 (720p) shows failed out of total 35 recorded. (Total is both 1080i and 720p)
Testing included 4 different 720p channels in single recordings, multiple and back to back configurations. No 1080i failed
Comskip failures are a known issue. I’m more concerned about “transport not connected” drive disconnects at the moment.
No problem, just trying to help track down reason for these failures, for when you have time to analyze.
One more thing … maybe a bit more descriptive.
There was only 1 720P show (which was a 1/2hr show) which had success on comskip out of the total 12 tested/recorded. The balance of 23 tests were 1080i, which were all successful. Multiple and back to back tests included a mix of each.
Thanks. Usually I can still get to the app area, either SSH’ing in with an app or a file browser. Worst case I could just root it if I really wanted to get to the files.
Did your Pro come with latest S/W or did you have to perform Update?
If latter, then a reset might help. See @karlmitchell comments.
For the past year using the Shield for Plex, I lost one or the other (or both) of 2 network storage locations on Shield multiple times, even though neither NAS disconnected from my network. Shield itself also seemed to maintain network connection (and I never did a power shutdown on Shield).
1 Like
IIRC each app’s files are completely private, so even with SSH or a file browser you can’t access them. The only option is to root the device, which is beyond the capabilities of most users.
Anyway, I think you can achieve what you want by
-
stopping the DVR (uncheck DVR box)
-
run this curl command:
curl --data-binary ‘{“Path”:“DVR”}’ -XPOST http://x.x.x.x:8089/dvr
This will end up storing your recordings in /data/data/com.getchannels.dvr/files/channels-dvr/data/DVR
I had to update, and I did a hard reset after I updated. Generally do that with any new device I get (and usually on version jump OS updates).
Done…
Let’s see how internal works 
Thanks a ton!