Film-Tech Forum ARCHIVE: GDC sync issue w/degraded array?

This is topic GDC sync issue w/degraded array? in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=002615

Posted by Scott Norwood (Member # 30) on 03-14-2016, 11:49 AM:

Has anyone else seen this? I did a screening recently-ish on a GDC server with a degraded RAID5. The sound sync was off by a few frames on anything that played off of the RAID (picture was ahead of sound). Playing DCPs off of CRU drives worked fine, with no sync issues.

This makes no sense to me, so I am curious to know if anyone else has seen this issue. Is there a fix or workaround? Does GDC have the ability to adjust sound sync internally like the Doremi does?

Posted by Steve Guttag (Member # 268) on 03-14-2016, 12:04 PM:

I have not seen it but yes, all servers we've come across can adjust lip sync on the server. For GDC it is in the "Configuration" section (can't be accessed while the show is running).

Posted by Frank Cox (Member # 6258) on 03-14-2016, 12:17 PM:

On the two occasions that I've had to replace the hard drives in my GDC server, there was suddenly a (consistently located) place in the movie where the picture and sound both started to stutter, just like the film was getting jammed in a real projector. "Blip blip blip blurp" and then it worked fine again until the end of the show.

Posted by John Roddy (Member # 7348) on 03-14-2016, 02:09 PM:

It shouldn't be possible. But as I'm sure roughly 100% of the people here will agree, that absolutely does not mean it won't happen. Equipment fails in seemingly impossible ways all the time. In this case, through no fault of GDC specifically, RAID failures can just make some really weird things happen. Luckily, GDC's software tends to acknowledge and adjust for that pretty fast. I've lost count of the number of RAID drives I've had to replace (we switched to GDC back in 2010, when Seagate drives were all the rage and nobody realized why they were suddenly so cheap). But every time it did encounter the error, a simple server restart was enough to permanently boot the failed drive from the RAID and run everything near-perfectly until a replacement set could be installed.

Diagnostics-wise, have you checked the RAID & drive info to make sure the bad drive actually was removed from the array? Again, I've seen the software do a pretty good job of automatically doling that, but there's nothing saying it won't.

Posted by Mark Gulbrandsen (Member # 72) on 03-14-2016, 07:36 PM:

If the server is over four years old then just install three new drives, I don't even waste time looking at SMART error logs any more unless the servers is still under its original factory warranty. This is especially true of servers that have been running on Seagate drives that were used while Hitachi drives were not. Most of those are now approaching four years of age. Drives are cheap, refunding shows is not. It's a no brainier. Every multiplex should have at least one set of spare drives in the parts cabinet! If a GDC kicks a drive out of the raid or shows a bad drive remove that drive from the raid. It will run just fine until you can install new ones.
And stick with using the correct HGST drive in the capacity you need!

Mark

Posted by Dennis Benjamin (Member # 1137) on 03-14-2016, 07:37 PM:

Yes, had it happen recently actually. We thought it might have been an amp, or a processor problem (skipping sound accompanied the out-of-syncness) - plus different issues occured at differrent times. It was indeed a degraded raid drive..

Posted by Nguyen Jack (Member # 7530) on 03-31-2016, 02:47 AM:

I think you should check the HDD (one by one) If 1 HDD problem, you can working with 2 HDD

Posted by Monte L Fullmer (Member # 2797) on 03-31-2016, 03:38 PM:

Had drive failure issues in my LMS were three drives lost it. We bypassed them in the config, reformatted the RAID and had to start over with re-ingests.