Nightly reboots: Why they must go...
- Thread starter John Kotches
- Start date
- Latest activity Latest activity:
- Replies 141
- Views 16K
But John saw no activity on his EHD(s) when watching his Hopper take 1 hr to reboot. So how could it be the EHD?
By the way John. How in heck do you get any fsck to complete in seconds on even a dinky disk? My multi-terabyte volumes take HOURS. Fortunately I haven't had to do an fsck in years!
By the way John. How in heck do you get any fsck to complete in seconds on even a dinky disk? My multi-terabyte volumes take HOURS. Fortunately I haven't had to do an fsck in years!
One thing NONE of us knows is if the Hopper actually NEEDS to reboot daily.
DISH has simply been making every model receiver reboot since the days of the 921 (which rebooted at 5 am) when they never could get that model to work right without a reboot daily.
DISH has simply been making every model receiver reboot since the days of the 921 (which rebooted at 5 am) when they never could get that model to work right without a reboot daily.
We're back to this being ext3fs which is journaled. So as long as the shutdown is clean there's no fsck that should be necessary on a daily basis.
I suppose we could discuss this, but not when I'm typing on Swype.
Sent from my Samsung Galaxy Note 2 using Tapatalk 2.x
What is the purpose of the reboot? I could see if new software was installed, but I never see anything change.
We don't know that it actually is fsck. All we know is that it is running a "System Integrity Check". What that entails is unknown at this point. We aren't following along on the console, are we?
Think about ext3fs, the journal and the check interval. Under most conditions you're using the journal. That is seconds to replay if that. With a clean shutdown it is already in sync, have a nice day.. When you hit the full interval you're running that no matter what.
My longest full fsck is about 28 minutes; for a Splunk volume of 3+ TB. We had to reboot for some tuning changes and had exceeded the check interval (180 days).
Sent from my Samsung Galaxy Note 2 using Tapatalk 2.x
It is a known fact that the head of Dish engineering reads these forums. And, he has responded to certain complaints, and corrected the issues. So, in this case, complaining on the Internet has fixed problems.
I agree that fsck is not required on a daily basis but I am sure of it that that is what is going on during the nightly reboots.
Sent from my iPad 2 using Forum Runner
That's a bit of a conundrum then. I was watching the drives and there was no activity yet you assert that it is fsck on the external drives that is slowing it down. If that's the case, where is the disk activity?
If it is fsck, it's the least active fsck I've ever seen!
Sent from my ASUS Transformer Pad TF700T using Tapatalk HD
I have no idea why your "activity lights" don't indicate activity but my EHDs have activity just about the entire time the "integrity check" AKA fsck is going on. Press YELLOW - BLUE - 3 - View Counters and scroll down to fsck and you will see that fsck is indeed going on just about every single night.
I don't get it either. If the firmware didn't update there should be no need for the reboot.
Unless they have something that's leaking memory (they don't have that much to work with) there is no reason. Even if they have a memory leak, unless it's the kernel, they should be able to restart the daemon and/or application that is running. That will release all the memory held by the program and let it start leaking again.
Sent from my ASUS Transformer Pad TF700T using Tapatalk HD
Read the DirecTV threads about their DVRs slowing down and needing manual reboot to restore speed. I think Dish's solution works better.
Then we will agree to disagree about the reboots. I'm coming from a position that servers, and the Hopper is a server, should be reliable and not rebooted on a routine basis.
Sent from my ASUS Transformer Pad TF700T using Tapatalk HD
As an IT manager myself, I can relate to John's argument. Dish should provide an option in the menu to turn off the nightly reboots. Perhaps they could make it a hidden feature like bridging currently is on the Hopper so the majority of subs keep it turned on.
I have a hard time understanding why Dish would knowingly release a product that requires a daily reboot. Seems to me that this would be a fix for beta testers and would be released to the general public once the issue was resolved. Making it worse, is the fact that it takes down every TV in your house (if you're relying on a Joey/Hopper solution). If it were my system, it would be kicked out of the house ASAP. My brother always complains about running out of tuners and this reboot. He is finally going to pay the early term fee and go back to DirecTV.
OR at the very least let us change the reboot time like the older receivers.
Sent from my iPad 2 using Forum Runner
Sounds like he needed another Hopper. I'd much rather deal with this reboot issue and have a super fast Hopper 95% of the time than DirecTV's slow receivers.
Sent from my iPad 2 using Forum Runner
The point is the mandatory reboot is not required to ensure "super fast" receivers. I manage servers that haven't been rebooted in months simply because it isn't necessary. How many of you reboot your computers daily? I'd bet very few do.
Never implied that that was the case. Just saying, Hopper is the fastest sat receiver on the market. I agree, especially with *NIX boxes, reboots aren't mandatory by any means with good programming.
Sent from my iPad 2 using Forum Runner
