Having a strange issue today. Noticed this morning a bunch of my recordings were showing drop% in the status, both HDHR and m3u sources. Figured it was me so I started troubleshooting.
My main CDVR is standalone, alternate is in Docker. Tried both, same drops. Swapped tuner priority around between my 4 HDHRs, same thing on all of them. Then I tuned an OTA channel directly on my client (bypassing CDVR entirely) and it played clean, no drops at all. Finally I rolled back a couple of pre-release versions thinking that it could be something related to that but the same issue happened.
So then I went into the CDVR logs and saw this on the latest tune:
2026/06/20 13:19:26.119786 [TNR] Opened connection to 104BD623/1 for ch49.1 ION
2026/06/20 13:19:26.120524 [DVR] Recording for job 1781975965-ch49.1 from 104BD623 ch49.1 into "TV/NCIS Los Angeles/NCIS Los Angeles S11E17 2020-03-08 Watch Over Me 2026-06-20-1319.mpg" for 41m4.190020133s
2026/06/20 13:19:26.175756 [DVR] Refreshing metadata for NCIS: Los Angeles (3561420)
2026/06/20 13:19:47.073832 [WRN] Timeout while getting series info for tms/3561420: retrying
2026/06/20 13:20:08.074343 [WRN] Timeout while getting series info for tms/3561420: retrying
2026/06/20 13:20:29.074866 [WRN] Timeout while getting series info for tms/3561420: retrying
2026/06/20 13:20:50.075321 [WRN] Timeout while getting series info for tms/3561420: retrying
2026/06/20 13:21:11.076935 [ERR] Failed to get series info for tms/3561420: Get "https://contentdb.fancybits.co/api/series/tms/3561420": net/http: request canceled (Client.Timeout exceeded while awaiting headers)
2026/06/20 13:21:11.146278 [IDX] Generating video index for job 1781975965-ch49.1
CDVR sat there for almost 2 minutes hammering contentdb.fancybits.co before giving up and starting the video index. During this time if I watch the status I see the buffer % grow then it fills to 100% and then the drop % starts incrementing. Thinking it could be DNS I tried changing to 8.8.8.8 but that yielded the same results.
Here's the weird part. If I restart the CDVR service, every rec that was showing drops comes back clean with zero drops and stays that way. But the second I layer a new recording on top, that new one starts dropping right away. The ones that were already running stay perfect.
My guess is the drops are happening at the very beginning of the tune while CDVR is stuck waiting on those contentdb timeouts. Once it gets past that, the rec runs fine for the rest of its window.
Anyone else seeing this or is it just me?
@maddox @eric @tmm1 I submitted logs in case that helps 6f661ca3-4fbb-48aa-a77b-60a562c07986
