1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

HR2x/R22 0x395: Is your Hard Drive loud?

Discussion in 'DIRECTV HD DVR/Receiver Discussion' started by Doug Brott, Jan 27, 2010.

Is your Hard Drive loud?

  1. Yes, I would describe it as loud

    0 vote(s)
    0.0%
  2. No, I would not describe it as loud

    116 vote(s)
    100.0%
  1. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,679
    348
    Dec 9, 2006
    Here's "my guess":
    When doubleplay was added, there was also a change to where the guide data was stored on the drive.
    It would seem this could be in the center of the drive and the buffering/recordings are on the outside.
    Since the DVR is always buffering, it's writing to the outside and then the guide data updates come and this needs to be written to the inside. This causes the head to "bang back and forth" as it is writing to both the inside & outside of the platter(s). Once the guide update is finished, it quiets back down and goes back to writing on the outer part of the platter(s).
     
  2. endeavor

    endeavor Cool Member

    44
    0
    Jan 27, 2010
    Antony,

    I only watch HD and that's what my testing is about. We have a 57" LED Samsung and am an Avid HD enthusiast. You can imagine why this glitch bothered us so much. (we don't watch SD very much but I've never seen it do it then no, but we all know the extra processing needed for HD compared to SD is substantial) I never have used OTA recording and it's not hooked up, I only use DTV satellite. The glitch happened on any HD channel they broadcasted, there was never one that I noticed that did it more than the other, they all did it.

    I read your other posts in this and the other thread too.

    Last night watching 3 more hours of live and prerecorded TV, I'm still happy to say the glitch has not happened even once in a total of now 10 hours watching for it, whereas before this latest software update and since last fall it would glitch a few times per hour - so far this is a Vast improvement ...but I'm still testing keeping an eye out.

    About this indexing noise the HD makes when it's doing this searching or cataloguing, I see it as the HD head going into overtime running like heck back and forth all over the HD's platters trying to get the info it needs for its outcome (while at the same time the HD is recording/rendering HD broadcasts) ...it seems so much effort for such a less important indexing function. I may not call or use the word thrashing which gives me the image of a fish being reeled in as it's splashing around, but it is certainly quite noticeable especially since the case of the DVR does seem to enhance it louder I agree. I notice the indexing search noise besides when the HR20-700 is off (but on I know) ...but also when watching TV and it did it's 'real-time' audio/video glitch and during that moment of speaker silence, well it was obvious hearing the indexing was in full swing (try using the Search function pressing keys at the same time the indexing was going on - it really took off even moreso then) In my observations, the 'real-time' glitch happened just as often with live tv or whether recording 1 or 2 simultaneous events - it didn't seem to matter, it just always did it a lot, and the glitch happened specifically while it was doing its indexing/cataloguing.

    The indexing noise certainly is still there after the update yes, but the important thing is that is does not 'so far seem' to interfere with HD playback like it fatally did before. Yes there is definitely still the normal pixelation/audio freeze (which is still a p.i.t.a.) which happens now and then and it is reproducible when you rewind and play back that segment it's still there, but we know that's from Network or Satellite broadcasting interference, and not resulting from real-time processes within the unit.

    HD's wearing out and going bad can make other mechanical noises yes for sure, but what I've been personally commenting on has nothing to do with that in any way.

    I still say going forward, that my advice to the DTV programmers is to code set this indexing/cataloguing process priority down even More, so it stands aside even more and gives 99% of processing priority to the recording/playback process where it's most important.

    Edit:
    Coming back to this post a week later to say with the 0x395 I've found that the issue is not completely fixed since it's done it to me thrice now within a week of a combined 12 hour period of watching HD programs, but that's much better than it was doing it many times per hour. I'd like to see this problem fixed even better ...wishful thinking I know....
     
  3. hdtvfan0001

    hdtvfan0001 Well-Known Member

    32,456
    258
    Jul 28, 2004
    I suspect that's an educated guess, and the likely correct explanation.
     
  4. endeavor

    endeavor Cool Member

    44
    0
    Jan 27, 2010
    Yes that sounds quite plausible for sure.

    Back in the last fall update they also added in all these other on-demand stuff, etc, to keep track of at that time too, and now with this extra cast/crew info tab there's even more stuff to keep track of. They need to update the processor in the newer units to keep up with efficiency processing power. Some day 10 years down the road flash drives will be big enough so that there will be no more spinning drives.

    edit: veryoldschool, thinking about it more, I would think though they would of had to format the HD to clear that center to place the guide info in the center, and then set it not movable, also format would delete stored videos, and so thinking about it I don't think any of that happened since I don't see DTV giving a command to format peoples HD's - that would create havoc. ..I'm just talkin' outloud, I'm not challenging any knowledge at all, just thinking outloud.
     
  5. HoosierBoy

    HoosierBoy Legend

    210
    5
    May 25, 2007
    Last summer, I had to have a tec replace my LNB and he moved me from a 5-LNB to the Slimline 3 LNB> My all of my hard drives were crunching since the update.

    LAst weekend, I had a H-20 go bad and the refurb unit did not work. I got a new H23 and the tech could not get it to activate. He discovered I was set up in the DTV system as having a slimline 5. When we called into the engineering group and changed me to a Slimline 3 in the system, the H23 activated quickly but all of my DVR's stopped the loud crunching sound.

    I was quite happy. Anyway, in case any of you can use this.
     
  6. TBlazer07

    TBlazer07 Grumpy Grampy

    4,785
    5
    Feb 5, 2009
    HUH? I don't believe the D* system has any record of which specific LNB you have on your roof and if it did it wouldn't affect your hard drive. Setting of which dish/LNB type you have is done on the receiver itself. It may have been set wrong and the tech fixed it on-site.

    The D* system MIGHT note if you are a SWIM customer vs. mswitch but that note is only if they need to send out a tech so that he brings the right equipment.

    Even with all that the dish type will only affect the programming you receive and even it if is wrong it would cause you not to receive ALL your programming or maybe even none at all. As for directly having an affect on your drive noise, frankly, I think that would be more of a coincidence. I'm no D* engineer but have actually used the D* "customer system" and had never seen anything referring to what dish or lnb you have. It's all receivers, programming and packages. Now it's been a year since I last saw it but I still know folks who do and that information simply isn't there to change.
     
  7. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,679
    348
    Dec 9, 2006
    You "might" want to change this to "does note" if you have a SWM. [I know it does]
     
  8. TalladegaTommy

    TalladegaTommy Cool Member

    20
    0
    Dec 27, 2008
    they just added fruit salad to a lousy dessert . how about lowering your prices
     
  9. hdtvfan0001

    hdtvfan0001 Well-Known Member

    32,456
    258
    Jul 28, 2004
    Using that example...if the food tastes that bad....one might want to change their diet. :D
     
  10. MattWarner

    MattWarner Legend

    265
    11
    Feb 11, 2007
    That certainly could be what is going on. The assumption I have been operating under these past few months is different..

    I believe it has to do with indexing the guide data. I think what we're hearing is the writing of the current channel buffer along with a index job that is running from time to time to index all the guide data (newly received indexed in with the existing). Remember: writing a buffer (or two buffers) involves writing large chunks to the drive. The head doesn't have to move locations as much. If you are writing large chunks AND doing large scale indexing (lots of small read/writes in a different part of the disc), I would assume the heads would be bouncing all over the place.

    These 'chatty' drives all started up one NR before the new search became active. My guess is that DirecTV did many of the under the hood changes for guide data (including indexing) one release before they released the new search.

    Anyway, my $0.02 worth...
     
  11. TBlazer07

    TBlazer07 Grumpy Grampy

    4,785
    5
    Feb 5, 2009
    I my case they had no record of it. Probably because I did my own upgrade. I had to email them to change it and they confirmed the change.
     
  12. TBlazer07

    TBlazer07 Grumpy Grampy

    4,785
    5
    Feb 5, 2009
    What does fruit salad and a diet have to do with a noisy hard drive? :)
     
  13. TBlazer07

    TBlazer07 Grumpy Grampy

    4,785
    5
    Feb 5, 2009
    I agree with you 100% on when it started and what caused it. In fact, that has been what I have been "preaching" since this started eons ago. Guide data, pictures etc are all part of it.
     
  14. rudeney

    rudeney Hall Of Fame

    4,266
    1
    May 28, 2007
    That sounds like a reasonable theory except for one thing (and this is a question, not an argument): Would the O/S (which is Linux) have the ability to dictate where on the platters any specific data is stored? I thought that the SATA interface pretty much hid physical disk characteristics from the O/S.
     
  15. hdtvfan0001

    hdtvfan0001 Well-Known Member

    32,456
    258
    Jul 28, 2004
    Both can give you a stomach ache? :D
     
  16. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,679
    348
    Dec 9, 2006
    I used "guess" for a reason.
    [another one here] the guide used to be on the outer edge of the platter(s) along with the [single] buffer. Then to add the space for the double buffer, the guide data needed to get move to another part of the platter that wasn't being used for recordings, since this space wasn't free for every drive/user. Imagine "all my recording went.... with this NR" :mad:.
    So since there is 100 GB of reserved "space" for DirecTV, they overrode some of "their space" with the guide data, and why it now is in the center.
    Since I haven't opened up my drive and watched what it is doing, this is all guess work from what "I sense" it is doing.

    I had a Seagate RAID with one drive being newer [cheaper] and this one drive was much noisier from the first. During writing to the RAID, it would sound like a machine gun, which is kind of like what my HR20-700 [seagate] does though not anywhere near as loud.

    So that's my story and I'm sticking with it, until somebody posts something better. :lol:
     
  17. Feb 1, 2010 #197 of 255
    bobcamp1

    bobcamp1 Icon

    896
    1
    Nov 8, 2007
    I think most of these people's hard drives are just fine as far as doing their basic function. I don't think they are dying. They are doing their job louder than normal. They do this by design when they are really stressed. And the HR20 enclosure is making things a lot worse.

    However, whether the hard drive is dying or not, that doesn't change the solution. If you have this problem, note that D* has had two software releases since it first occurred. Neither of them fixed the issue. If you think the next release will fix it, then go ahead and wait. I don't think it's going to get better any time soon if ever. If the noise is really annoying, have D* replace your HR20 with another HR model. That's one of the benefits you have with the lease model. Take advantage of it.

    Maybe if D* gets enough HR20 returns in, someone over there will notice a pattern and try harder to fix it.

    If you own the unit, get an Antec MX1 eSata enclosure and transfer the hard drive from inside the HR20 to this unit. (Note: you'll also have to put a "dummy" hard drive inside the unit so it will boot properly.) They are the enclosure of choice for S3 Tivos, and we (thanks to TBlazer07) have evidence that they (unsurprisingly) are just as good for HR20s for all the same reasons.
     
  18. Feb 1, 2010 #198 of 255
    Doug Brott

    Doug Brott Lifetime Achiever DBSTalk Club

    28,939
    72
    Jul 12, 2006
    Los Angeles
    I think there may be some real merit to this line of thinking ...
     
  19. Feb 1, 2010 #199 of 255
    bobcamp1

    bobcamp1 Icon

    896
    1
    Nov 8, 2007
    The answer to that is no, but mainly yes!

    No: the hard drive always presents a virtual map of sectors to whatever it's hooked up to. Two "adjacent" sectors may be on opposite ends of the platter, especially if a bad sector has been silently replaced with one of the spare good ones. Also, the OS typically doesn't have access to the "System area/negative/reserved cylinders" used to store the hard drive's firmware. This link is a pretty good summary for more info.: http://www.myharddrivedied.com/presentations_whitepaper.html

    Yes: in general, the mapping that the hard drive shows to the O/S does correspond with areas of the hard drive. The O/S (or installer) is responsible for setting up the partitions. The O/S also decides where to place files within that partition and which files go into which partition.

    Although defragmented files and multiple partitions allow for slightly quicker loading, they can cause problems with multi-tasking O/Ss. Each task may be trying to access a different file at the same time. If those two files are frequently accessed together, but are spaced far apart on the hard drive, you're going to get the behavior we've been seeing with the HRs. In theory, these files could be interleaved to reduce the problem.

    However, most O/Ss do not defragment on the fly. Almost none of them interleave files intentionally. Especially Linux. I don't know what D* did, and even if they did it intentionally or not. So it's tough to figure out if they can fix it.
     
  20. Feb 1, 2010 #200 of 255
    TBlazer07

    TBlazer07 Grumpy Grampy

    4,785
    5
    Feb 5, 2009
    That won't work. Without an internal drive the HR won't boot. I am told that it used to but both my HR20-100 and HR20-700 won't when I remove power from the internal drive under my current firmware.

    Best way is just a new external or take the internal out and make it external and put a "dummy" drive internally. That way you'll still have your programming.
     

Share This Page