Film-Tech Forum ARCHIVE: GDC SA2000 not playing DCPs, running very 'choppy'

This is topic GDC SA2000 not playing DCPs, running very 'choppy' 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=002535

Posted by Ethan Dutcher (Member # 8319) on 12-30-2015, 11:44 AM:

I just restarted my GDC server trying to set-up some captioning glasses we just got. When I started the DCP after restarting I noticed the slider moving very 'choppy', not smoothly like it normally does. It is also taking much longer than it should to move forward. But the worst part is that DCPs are not playing at all. I obviously removed the glasses transmitter immediately and restarted the server and I am still getting the same issue. I did a complete power down and back up and it is still happening. I'm trying to call GDC right now, but has anyone encountered this before? I have a show in 6 hours so I'm obviously very concerned right now.

Posted by Phil Ranucci (Member # 3820) on 12-30-2015, 12:04 PM:

Long shot-make sure that the drive locks are actually locked and the drives aren't a bit loose in the frame. We had somebody with big fingers unlatch one trying to remove a CRU drive.
All DCPs or just one?
Did you try deleting the DCP and reloading?
Try live playing the drive if you still have it.

Posted by Ethan Dutcher (Member # 8319) on 12-30-2015, 12:43 PM:

Got through to GDC and they had me re-seat the IMB card in the projector and PCI-e cable and everything is working again. Didn't even need to re-marry the projector. The tech couldn't give me an explanation why it happened though. I'm concerned since it happened after a restart there might be a problem with the server start-up procedure. I ran a system check while it was happening and everything came back OK. Guess I'll just keep and eye on it. I'm the only one who restarts the system unless there is a power outage anyway.

Posted by Mike Blakesley (Member # 26) on 12-30-2015, 12:44 PM:

We had a similar issue with choppy playback and it turned out to be one of the hard drives. If you can identify which drive it is, you should be able to pull that drive and it'll still work with the other two.

Posted by Mark Gulbrandsen (Member # 72) on 12-30-2015, 12:49 PM:

Yea, go to control panel and see if all your hard drives are present and have GDC check the SMART errors on each drive. Skipping is the most common issue when hard drives begin to fail or you lose sectors on a drive. Hard drives are cheap and considered an expendable item, so if in doubt just replace them. Especially is your unit is more then 4 to 5 years old.

Mark

Posted by Ethan Dutcher (Member # 8319) on 12-30-2015, 01:54 PM:

I'm getting OK's back from all drives and the RAID check came back fine. I did just replace one of the hard drives over the summer, so I know that they are getting to the end of their life. I do have a spare on hard, but without knowing which one is having the issue I don't know which to replace. The server is getting to about 5 years now, Is it worth just replacing all the drives and rebuilding the whole RAID?

Posted by Mike Blakesley (Member # 26) on 12-30-2015, 02:45 PM:

Yes, it is. But if you need to run shows ASAP, try taking the drives out one at a time, rebooting and see if you can playback smoothly. That way you can eliminate the bad drive. Then at your convenience, replace all of them at once.

Posted by Steve Guttag (Member # 268) on 12-30-2015, 04:21 PM:

IF you have admin access...you can check the drives yourself by going to the Storage System management. It will allow you to have it read each SMART report (one at a time) and look for item 5 (Reallocated Sectors). If it is over 100, change it. Make a note of where they are now and check a day later...if they increase in just a day...change it.

Also look towards the bottom of the report and look for ANY errors...if it has them...change it! As Mark said, drives are cheap, muffed shows are expensive and annoying.

Posted by Monte L Fullmer (Member # 2797) on 12-30-2015, 05:48 PM:

Does sound like a flaky HD in the RAID

Posted by Dennis Benjamin (Member # 1137) on 01-01-2016, 10:06 AM:

We just had 2 hard drives fail in a raid at the same time on a GDC server.

I ran our feature as "live play" until G.D.C. was able to get us the new drives.

All is good now...

Posted by Steve Guttag (Member # 268) on 01-01-2016, 10:16 AM:

Here is the thing...did they REALLY fail at the same time or did one of them fail or effectively fail (internal errors that nobody noticed because it is a RAID 5) and when the other drive failed, the RAID crashed.

Out of curiosity, which brand drives were they and how old were they? The other thing I'm doing with our GDCs is if we are going to change more than one drive, swapping them all out for 2TB drives to at least get 4TB of storage on them. The 2TB is a bit of a source for complaints. GDC will sell the SX-2001A, SX-2000AR with 4TB capacity now. The SX3000, via the Enterprise/Enterprise + can now have quite a bit of storage, finally (and with 3.5" REAL drives). I don't know about others, we get constant requests to go beyond 3TB of capacity with 6TB seemingly comfortable, for now.

Posted by Mark Gulbrandsen (Member # 72) on 01-02-2016, 01:11 PM:

I've had that dual drive failure in both the DSS-100 and DSS-200, but only once in each. Never had two drives fail in anything else, ever.

I only have one second run site that has 4 TB of storage. No one even comes close to needing that. First run shouldn't really require that unless a manager likes to move shows around instantly and they all have to be loaded in every server...

Mark

Posted by Carsten Kurz (Member # 5396) on 01-02-2016, 04:57 PM:

Hmm, interesting thing about the live play function - that's a really helpful option if it works with a blown raid on GDCs.

Posted by Steve Guttag (Member # 268) on 01-02-2016, 07:56 PM:

Can one do a live-play on a 3000?

Again, ALL of our customers, multiplex/art house...you name it want MORE STORAGE and more than 3TB. Our DSS systems will start to migrate to 6TB over time and other brands will migrate to 4TB or more.

Posted by Mike Blakesley (Member # 26) on 01-02-2016, 10:59 PM:

quote: Mark Gulbrandsen
First run shouldn't really require that unless a manager likes to move shows around instantly

I'd bet that's what they ALL want.

Posted by Ian Freer (Member # 1940) on 01-03-2016, 12:06 AM:

quote: Steve Guttag
Can one do a live-play on a 3000?

Steve, yes you can Live-Play with an SX-3000. You of course need a CRU RAX unit (or a Move-Dock connected via eSata)...

I have not installed any Enterprise Plus units yet to know if you can Live-Play from the built-in CRU slot, but I find my customers feel the same as yours, they all want more storage, so I'm looking to upgrade some from the PSD to the Enterprise or Enterprise Plus over time...
My largest Enterprise units are currently running with 5x 2TB drives for approx ~7.5 TB storage, but now GDC is supporting 4TB drives so a ~15TB RAID is an option.

Posted by Mark Gulbrandsen (Member # 72) on 01-03-2016, 11:08 AM:

The very latest SX-3000 firmware upgrade (Build 383) now allows eSATA ingest and live play with the latest version of the move dock 3S. Up until now that version would not work because the eSATA ports on the SX-3000 were not port multiplier aware. The firmware upgrade corrects that issue. You had to have an older move dock 2, or the more expensive RTX version dock to ingest via esata. Those older CRU units have completely different chipsets in them than the 3S does. The Movedock 2s has been gone for over two years now...

I discovered this issue when I installed a new SX-3000 at the Roxy Theater and the 3S couldn't be detected by the server. We worked with CRU and GDC to rectify this. GDC Tech support was not aware nor had a Move pocket 3s been tested by either CRU or GDC. It's always fun to be the first! Took a while for the fix but it's not an issue any more. Note that this is apparently an issue that also exists with the IMS type servers as well. But what issues don't those have....

Mark

Posted by Dennis Benjamin (Member # 1137) on 01-04-2016, 07:50 AM:

quote: Steve Guttag
Here is the thing...did they REALLY fail at the same time or did one of them fail or effectively fail (internal errors that nobody noticed because it is a RAID 5) and when the other drive failed, the RAID crashed.

Probably what you are saying, one drive failed, no one noticed, then the other one failed.....

quote: Steve Guttag
Out of curiosity, which brand drives were they and how old were they?

Hitachi, 2012 - the ones that came with the server...

Posted by Steve Guttag (Member # 268) on 01-04-2016, 08:01 AM:

Thanks.

Posted by Ethan Dutcher (Member # 8319) on 01-04-2016, 09:30 AM:

So, the hard drive tests are telling me all the drives are going bad(everything is labeled old_age or pre_fail). Even the one I replaced in June which seems strange. Just want to make sure, creating a new RAID is going to take about as long as rebuilding, 6-8 hours, right? I want to ensure I have enough time to reingest and test everything afterward.

Posted by Ethan Dutcher (Member # 8319) on 01-04-2016, 11:12 AM:

Also, the server natively supports drives over 1TB right? That wasn't a firmware/SMS upgrade? I'd like to put 2TB drives in there since they are fairly similarly priced and we could use the extra storage.

Sorry for all the questions, I'd read the manual but all I can find are the user and install manuals which don't cover any of this.

Posted by Mark Gulbrandsen (Member # 72) on 01-04-2016, 01:34 PM:

Mike,

Ethan..... It's still best to have the latest version firmware in your server. Just call GDC and ask them....

The only location that I work on that does that is a 2nd run theater in Ogden, they do have 4TB raids there. The rest of the locations pretty much know which screen to open a film on and when to move it to a smaller screen. Note that this move can even be scheduled on the TMS to automatically happen or the manager can do it remotely. Note too that GDC servers, even the IMB versions (not the SX-3000) can be powered up remotely WITHOUT the projector having to be on in order to move the content via the TMS. For the SX-3000 you just remotely power up the projector.

Mark

Posted by Ethan Dutcher (Member # 8319) on 01-04-2016, 02:00 PM:

Thanks Mark, I know it is best to have the most current firmware, but since we are out of warranty with GDC I am terrified of a firmware update bricking the server and GDC saying they won't help. With things currently mostly working I'm hoping to get through until the server is replaced in our planned expansion.

We don't have a TMS(one screen currently) and tend to play 3-5 different DCPs a week here, the 1TB drives work most of the time but I have run into points where the server has gotten uncomfortably full. We also do a festival once a year and tend to fill up the server that week as well. I guess since we are planning to remove the server soon enough it probably isn't worth putting in the extra space since we are able to get by.

Posted by Mark Gulbrandsen (Member # 72) on 01-04-2016, 02:27 PM:

I never had that happen except at very old firmware versions and it won't brick the server anyway. The update does not load in until after it is finished ingesting it and it has been verified. If it doesn't load correctly you just do a cold boot and then start over again. If you are at some older version you would have to upgrade in stages according the the range the firmware allows, or just buy a new OS hard drive which should have the latest version on it. Not knowing what version you are at it might matter when it comes to playing back SMPTE DCP's. GDC is also selling extended warranties very reasonably these days and would likely do a one time update charge and do it remotely for you.

Mark

Posted by Lance C. McFetridge (Member # 95) on 01-04-2016, 03:33 PM:

Ethan:
Be sure you understand how to interpret the SMART diagnostics under the admin. Steve is correct with looking at line 5 as well as line 196 and 197 of the report. What you are seeing may not indicate a failing drive. If you can access the admin portion of the GDC software, you can remove the drive in the raid if it indicates failure. Live play is a good solution but the GDC software looks to the raid first then live play. So remove any file you are live playing from the raid then mount the CRU.
Lance