dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
6321
share rss forum feed


beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
reply to nunya

Re: CATV issue- box freezing and reboots.

said by nunya:

You might consider one of these: »www.cabletvamps.com/Products/ERA-4100.htm to use in conjunction with your existing amp.

Do not install this. Reverse amps should only be installed by the MSO, not a homeowner without a meter. They are very rarely needed. From the sounds of it the issue is most likely outside of your house.

Have you tested a single box bypassing the amp yet?
--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
said by beachintech:

said by nunya:

You might consider one of these: »www.cabletvamps.com/Products/ERA-4100.htm to use in conjunction with your existing amp.

Do not install this. Reverse amps should only be installed by the MSO, not a homeowner without a meter. They are very rarely needed. From the sounds of it the issue is most likely outside of your house.

Have you tested a single box bypassing the amp yet?

I offered to do that for the tech, but he is convinced that it's not the amp nor the house wiring, and declined my offer to do that. He is suspecting it's something in the headend or a return signal issue. They're hoping to catch a freeze-up on the sling box and take it from there.

When he and I checked his laptop displaying the noise on the return frequency, it looked pretty clean (at least that moment). Is my typical return signal strength of 34-35 dBmV high enough to overcome most noise? I know a cable modem can go upwards of 55 if needed to overcome noise... what should I be pushing to get a quality return signal?


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to beachintech
said by beachintech:

Ingress that seems to be taking out that QAM possibly. Again, without a meter that's hard to diagnose over the web.

98% of the time, I can simply tune to a switch digital channel or get On Demand programming with no issues... it's very weird and random when this freezing happens. I had 2 freeze-ups in the last 24 hours and documented the times. I'm hoping the sling box caught them so that a diagnosis can be made.

If it were ingress, wouldn't I be having lots more problems, lots more often... or others be in the same boat as me? Someone watching popular channels at peak times wouldn't notice i suppose.. but someone watching Sprout in the middle of the night might.

But again, if it was noise, why can I simply move the channel and return to refresh the signal, or just tune to a channel and immediately get the feed 98% of the time... shouldn't that be problematic at least a decent amount of the time?!?

Seems to me that the initial request is making it though but the "pinging" of the STB back to the node to keep the feed going is what's not working... but why that and not the initial requests?!?


beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
There are a lot of things that can cause intermittent problems, someone keying up a mic, etc. We had an issue like that here, it took a long time to resolve, but it turned out to be a bad piece of equipment in the field (Amp AGC). This was VOD though, we didnt do SDV when I was there.
--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


cypherstream
Premium,MVM
join:2004-12-02
Reading, PA
kudos:3
reply to SwedishRider
Did they do a software update recently to the SDV client or set top boxes?

Sounds like when you change the channel the SDV request is made, and delivered. But the software in the box is not generating the heartbeat and it times out eventually. Or it does generate a heartbeat but after 360 minutes the SDV client does not inform the UI to display "If you want to continue watching this channel push A" (or whatever the message is). The timeout value on the SDV gear is configurable and the timeout value on the guide software is also configurable.

Perhaps there was a software update or firmware update to SDV servers that reset the timeout to the default 360 minutes, but the client guide software in the set top boxes are set to something like 600 minutes. Therefore the "press this key to continue watching" prompt never makes it before the SDV session is torn down. The last keyframe seen is what ends up being frozen. Change the channel forward and back and the STB OS sends the request to build the SDV session again.

The session is only torn down if no one else in the node is watching the channel within the timeout value. So it may not happen on a more popular network than say Sprout, which is usually on a certain level of service and not watched as frequently as something more mainstream. This contributes to the lack of reports on the issue and why its taking so long to figure out. Try another channel that is low on the totem pole for your area. Like Ovation or Reelz channel, or maybe Horse Racing TV if you have that. Anything that's not on a 'basic' tier.

Just a thought. There's a lot of pieces at play software and IP networking wise between session resource managers, sdv, oob carousel, DAC (moto) or DNCS (Cisco/SA), and edgeqam's.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
said by cypherstream:

Did they do a software update recently to the SDV client or set top boxes?

They've done firmware updates on and off, but this has been going on for years to be truthful. I have not been pushing the issue until now since my contract is up and I'm free to shop around, but I've had multiple CISCO/SA boxes swapped in and out... same thing

Sounds like when you change the channel the SDV request is made, and delivered. But the software in the box is not generating the heartbeat and it times out eventually. Or it does generate a heartbeat but after 360 minutes the SDV client does not inform the UI to display "If you want to continue watching this channel push A" (or whatever the message is). The timeout value on the SDV gear is configurable and the timeout value on the guide software is also configurable.

The STB has and does generate the "press this key to continue" message before timing out after hours and hours of use, but the freeze-ups are randomly happening much sooner than 360 minutes. It could be 45 minutes after switching to a channel. It could be 3.5 hours... but most of the time, it's pretty problem free. I think it's more of the "heartbeat" not being sent or not getting through and the node shutting off the stream.

Perhaps there was a software update or firmware update to SDV servers that reset the timeout to the default 360 minutes, but the client guide software in the set top boxes are set to something like 600 minutes. Therefore the "press this key to continue watching" prompt never makes it before the SDV session is torn down. The last keyframe seen is what ends up being frozen. Change the channel forward and back and the STB OS sends the request to build the SDV session again.

The session is only torn down if no one else in the node is watching the channel within the timeout value. So it may not happen on a more popular network than say Sprout, which is usually on a certain level of service and not watched as frequently as something more mainstream. This contributes to the lack of reports on the issue and why its taking so long to figure out. Try another channel that is low on the totem pole for your area. Like Ovation or Reelz channel, or maybe Horse Racing TV if you have that. Anything that's not on a 'basic' tier.

It's not after 360 minutes though... it can be much sooner than that. But I agree it's with less popular channels. Typical freeze-ups I posted of were on Sprout, NBC HD, even the Smooth Jazz music channel has frozen... so I agree those are probably not being watched anywhere else on the node.. but why is the return signal not being generated or getting through to the headend, but a quick channel change will fix the problem?

Just a thought. There's a lot of pieces at play software and IP networking wise between session resource managers, sdv, oob carousel, DAC (moto) or DNCS (Cisco/SA), and edgeqam's.

Yeah, I feel like it's needle in a haystack time. I'm hopeful it can be fixed relatively quickly... but if not, I'll have to decide if I just want to "live with it" or keep internet and phone (which are working flawlessly) and resubscribe to DIRECTV. I'd rather keep my bundle with Charter... but this doesn't seem to be an easy fix, so I don't know how long they'll chase this until they decide it's too random to diagnose. They have really been on this so far, so I'm hopeful they catch it and get it fixed.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to SwedishRider
Techs came back for another round today, and here was the deal:

Slingbox was clean, so the problem is definitely between me and the node.

Most likely suspect at the moment: low signal strength on the return path, so they eliminated the amp and went with a 8-port splitter, which has a signal loss per port of 11 dB I believe. That resulted in the return signal to jump to about 46dB per box... but....

It also reduced the forward signal strength, and some channnels now report values of -8dBmV to -10dBmV.

For the moment, all TVs work, and this will help to diagnose if the issue truly is weak return signal. I'll have to let some time go by freeze free before I can determine if this is the answer to the freezing problem.

My concern, though, is the forward power signal swings due to the weather. It's cold out there now... things will change in the summer. What are the realistic tolerances of a CATV STB with regards to signal strength and still getting a quality picture? Is the range -10 to 10, or wider than that? Will I have enough room to get some swings and not get audio issues or macroblocking?

Tech said if this fixes the freezing issue, but problems arise on picture quality... then it's time to hook the amp back up and pad each box based on what it needs.

Thoughts?


nunya
LXI 483
Premium,MVM
join:2000-12-23
O Fallon, MO
kudos:12
See the link I provided above.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to SwedishRider
Well, RDC is showing 45-47 dBmV on 19 MHz, so I tuned a box to a channel that I thought would be unpopular in the afternoon (obscure movie on a premium channel) and waited to see what would happen...

After 30-45 minutes... it froze.

I'm going to give it the weekend and see if it happens again... but if return signal strength is ruled out... what is left at this point?


Killa200
Premium
join:2005-12-02
Southeast TN
Reviews:
·Charter
reply to SwedishRider
This is sounding a lot like low band sporadic ingress. If ingress was causing excessive return path noise at the time the heartbeat was being sent from the STB, causing the singal not to make it to the SDV setup at the headend, it would collapse the channel from no response. I'm not sure what the channel timeout is after failing to receive heartbeats, so I'm not sure how bursty the ingress would need to be. I can say though that while there is no choice in it due to how narrow the band is, it sucks to be running return at 19mhz.

Cable modem could sustain a certain amount of bursty ingress, especially since it is more than likely higher in the return band where it is cleaner. Even though data / voice is working, it would be interesting to see the T3 / T4 modem errors to see what the plant is telling us.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
Cable modem is reporting upstream on 30610000 Hz at 40 dBmV. Logs are all fouled up from plugging and unplugging everything over the last week.

But yes, typical uptime has been 20-40 days for the cable modem. And it's only the heartbeat that is not working... a simple channel switch and then return sends the request, which restarts the stream... I'm really not having problems with requests... just keeping the session alive.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to Killa200
Yep... it's confirmed. Problem is still happening.

Since I'm typically an insomniac, I tuned a box to a premium channel in the middle of the night with an old movie on that I suspected would be unpopular... and sure enough, within an hour, channel froze up. A quick channel change and back, and all was right with the movie. WTF!?!?!

It's the middle of the night, my house is virtually powered down- nothing is on, neighborhood is silent as well... and time of day seems not to matter. And it's still all three boxes doing it... cable length makes no difference. I just don't get it-this isn't making sense.

CISCO glitch somewhere? Could it be my boxes are just not configured to send out the heartbeat, but configured to start the initial session?!? Again, initial requests are clearly getting through... changing channels immediately starts the channel I'm tuning to, and at 46 dBmV, the headend has to be hearing me (otherwise channel changes on switch digital would not be working).


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1

1 edit
reply to SwedishRider
It just happened again, but I think I can replicate this now... session times out at either 59 or 60 (not exactly sure) minutes from the moment I select the channel (but obviously that assumes that nobody else in my node is tuned into the channel to keep the session alive). It was about an hour last time, and it's exactly 59 or 60 minutes this time.

Hmmm......


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to SwedishRider
Yes, I can replicate this... it happened yet again on another obscure channel/movie after 1 hour... a pattern doth emerge...

Methinks this is a setting in the boxes that isn't set to properly send the heartbeat packets, or a setting at the node that is not recognizing the heartbeat packets. It's not a return path issue, it's an actual setting issue.

Agree? Disagree? Besides calling this in to Charter (I have local's number to the Tech supervisor), anything else I should do or consider?


beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
The node doesn't care what's hitting it, all it does is convert light to RF and RF to Light on the way back. They have a trunk issue / HE issue they need to resolve for you. No return amp is going to fix that, no matter what the sparkys say.

I hope they get it resolved, but you are at their mercy until they figure it out. I suggest to keep calling until it's fixed. Squeaky wheel gets the grease.

Have you tried hooking 1 (one) box up to the two way where the modem is to replicate yet, with everything else in the house disco'd?

--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


nunya
LXI 483
Premium,MVM
join:2000-12-23
O Fallon, MO
kudos:12
Reviews:
·Charter
·voip.ms
·surpasshosting
reply to SwedishRider
Sounds like a software glitch to me.
God be with you. It's tough dealing with cable guys. They usually aren't the sharpest knives in the drawer to begin with. I know, I used to be one.

How many TV's do you have? If Charter can't resolve it, you may want to look into Direct or Dish for your video purposes.
One of the only reasons I stuck it out with Charter so long last time was we have 8 TV's. I finally went back to OTA with MythTV, PlayON, Wii, Xbox, and Roku boxes
--
If someone refers to herself / himself as a "guru", they probably aren't.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to SwedishRider
Here's my thesis... this is edited from a post in Charter Direct:

Hi Abby,

Sorry for the late reply. I haven't personally spoken to XXXX as of yet, but I have left him a few voicemails and he's either dispatched a tech with info or had the tech call me. This is turning out to be much more difficult, but after last night, I think we've narrowed it down again.

Charter techs changed out my amp for a traditional splitter to drive up the return decibel level. The thought was that the head end couldn't "hear" my set top boxes, and so was ending the switch digital sessions because the headend was missing the signals returning from the settop boxes/DVRs. At the time, the freezing was so intermittent, it was the only logical thing that made sense.

Last night, freezing problems reappeared even with those corrections. However, being the insomniac that I am, I did another test. I picked a premium switch digital channel with an old movie late at night that I thought nobody else in my node was watching and let it run.

Results? Channel froze after about an hour..

So I picked another movie on a premium channel late at night (on Encore I think) and let it run again..

Results? Channel froze after about an hour..

So I went to a different cable box in the house, still early in the morning, and did the same thing... picked an old, not-so-popular movie on a premium channel and let it run.. You guessed it... it froze after about an hour.

I can now replicate the problem. What is apparently happening is NOT that the boxes can't communicate back to the head end, it's that they AREN'T communicating (at least properly) with the headend. This isn't a wiring issue on the pole or my house... This is a settings/software issue on the node. The only time the command to open a session is initialized is when a channel is first tuned by the user... but afterwards, the box isn't sending the "keep-alive" or "heartbeat" signal... it is only sending the initial request to open a session. That's why my remote can fix the freeze by changing the channel and returning. It always fixes the problem.

The reason this has been so intermittent is because if I am on a switch digital channel, even if one other cable box in my node tunes into the frequency as I'm watching, it will reset the 60 minute window... making it unpredictable when freezing will happen. To someone watching only popular channels during busy times.. freezing issues might never happen. But for me, who uses non-traditional channels quite a bit for longer periods of time, this is a problem, as my session will end because I'm the only one watching and nobody else is tuning in to keep the channel session alive. For football fans.. this won't be an issue... but for Sprout TV users... it's a big issue.

My ideas:

1) Set top boxes/DVRs are not set to properly send the "heartbeat" or "keep-alive" signal before the switch digital channel times out (seeming to be 60 minutes), thus ending the session, and giving me a frozen channel.

2) Node/Headend is not set to properly receive the "heartbeat" or "keep-alive" signals before the switch digital channel times out, thus ending the session and giving me a frozen channel.

3) Timing issue... the set top box is set to send out the heartbeat timed after the node is programmed to close the session, resulting in a frozen channel after 60 minutes.

4) CISCO/SA glitch, which would require a help ticket with CISCO (really hoping that's not the case).

The reason this has been so hard to diagnose is that the problem has been intermittent... and for good reason. If even one other person in my node tunes to the channel within that 60 minute window, the session will honor that request and I will continue to get the signal past the 60 minutes. For people who watch popular channels... this may never appear, or appear so infrequently, it's not worth mentioning... but users of non-typical channels for over 60 minutes keep getting frozen... and that's where I am. As I said before, if I was a football fan... this issue might never have come to light.

I don't claim to be a cable guru, but I'm pretty confident in what I've posted above as we've ruled out most everyting else, and my test last night over 3 times yielded the same results. Replicating the problem has hopefully pinpointed the issue, or at least given an exact place to look for problems. XXXX and his techs have been all over this thus far... I do have confidence that they can get this one nagging issue ironed out.

If you could send this posting by email to XXXX, I'd be appreciative. I also left him a voicemail, but this is more detailed and sometimes an email just works better. My cell is XXX-XXX-XXXX. If he could call me directly when he gets a minute to chat about this, I'd be appreciative.

And as a final note... I did read about Charter's decision to close their social media department and end their affiliation with DSLR. I am saddened and somewhat angered at the move... you have provided excellent first response service in these forums, and that will be sorely lacking going forward. I'm not sure if you've decided to take a position with Charter phone at the end of the year... but in any case I wish you and the entire Charter Direct team all the best... you will be missed here in the forums.

Please let me know when you've passed this along to XXXX... and please have a very happy holiday season!

Thanks!
Jim



SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to beachintech
said by beachintech:

Have you tried hooking 1 (one) box up to the two way where the modem is to replicate yet, with everything else in the house disco'd?

I haven't, and I'm not sure I need to at this point. Clean signal is making it back to the node- every channel change works to start a session and to fix a freeze, regardless of box, outlet, etc. Now that I can replicate it after a consistent 60 minute period... I think we can rule out ingress.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to nunya
said by nunya:

How many TV's do you have? If Charter can't resolve it, you may want to look into Direct or Dish for your video purposes.
One of the only reasons I stuck it out with Charter so long last time was we have 8 TV's. I finally went back to OTA with MythTV, PlayON, Wii, Xbox, and Roku boxes

I have too many TVs

I'm not sure what I'm going to do. In fairness, Charter's techs have been awesome tracking this down.. so I'd like to believe they will see this to correction. If not... I'll take that if/when it happens.

Honestly, I've liked Charter except for this nagging issue... I'd like to stick with them for all my services if possible.


beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
reply to SwedishRider
Depends on what freq it's on, but possibly. They are going to ask you to do it at some point.
--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


Killa200
Premium
join:2005-12-02
Southeast TN
reply to SwedishRider
To me it sounds less likely to be a box issue because your saying it happens over multiple brands and models of boxes, which means that would cover a wide range of different firmwares, both make and revision.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
said by Killa200:

To me it sounds less likely to be a box issue because your saying it happens over multiple brands and models of boxes, which means that would cover a wide range of different firmwares, both make and revision.

Good point. I never thought of that. All three boxes are different- one is a SA HD DVR, one is SA non-HD set-top, and one is CISCO HD DVR. All of them exhibit the same behavior.. so it's more likely to be a problem with the node or headend than in the boxes themselves...

Unless Charter never configured the heartbeat in any of their boxes?!?! That seems pretty unlikely though...


nunya
LXI 483
Premium,MVM
join:2000-12-23
O Fallon, MO
kudos:12
Cisco = Scientific Atlanta.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
said by nunya:

Cisco = Scientific Atlanta.

I did know that... but isn't killa right about different firmware for each box?


nunya
LXI 483
Premium,MVM
join:2000-12-23
O Fallon, MO
kudos:12
It's probably not that different, and they all talk to the same node and headend.
--
If someone refers to herself / himself as a "guru", they probably aren't.


beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
reply to SwedishRider
Yes, but they have the same API's to interface with the headend equipment. Nothing in your house talks to a node, it's a device doing light conversion.

Yes, Cisco bought SA some time ago, new equipment will carry the Cisco logo. They generally use the same firmware, with some exceptions.

You most definitely have either a trunk issues with the amps in the field or they have a headend problem. If it was a heartbeat issue related to firmware no one would be working, which would be a huge problem that would have been noticed prior to your calls.
--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


nunya
LXI 483
Premium,MVM
join:2000-12-23
O Fallon, MO
kudos:12
Reviews:
·Charter
·voip.ms
·surpasshosting

Nothing in your house talks to a node, it's a device doing light conversion.

Actually in newer SDV systems, node groups use a different RPD than the "@ node" resource manager. E.G. the headend doesn't give a rats ass or need to be bothered if another subscriber is added to the stream locally.
It's not simply a repeating device as you allude. It's a separate ARPD, and therefore vulnerable to failure (hardware or software).

Jim - have you talked to any of your neighbors?
--
If someone refers to herself / himself as a "guru", they probably aren't.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
said by nunya:

Jim - have you talked to any of your neighbors?

Interesting that you asked...

I previously called my neighbor (who has charter for CATV) when I started investigating this, and asked him if he has the same symptoms as me. He said no.

Today, however, the tech came back to hook my amp back up, as my signals levels were scraping the bottom and causing some macroblocking when they eliminated the amp in favor of a simple 8-outlet splitter (and did nothing to fix the freezing issue). As the tech was heading down my driveway, my neighbor caught me... his complaint to the tech? He said his TV froze, and a simple channel change fixed the channel. When asked, my neighbor said he is a serial "channel jumper", so he always bounced around and never saw this problem before. A relative is staying with them and is a consistent channel watcher... and they are now getting frozen channels!

At this point, tech said it's confirmed to be a node/headend issue of some kind. At least 2 homes in my service group are reporting the same symptoms (my neighbor and me)... it's not my wiring, it's a service group/node/headend issue of some kind.

And to complicate things, I thought that the problem was a 60ish-minute consistent timeout window, but my wife tuned to Sprout today... it froze within 8-10 minutes, and was fixed by a simple channel change and return. I'm still hoping it's a simple settings issue at the headend... but I'm starting to fear it's going to be a needle in a haystack situation.

Tech said that to his knowledge, Charter's local SDV system does not rely on a heartbeat... it relies on a 360 minute (6-hour) window for the channel to expire it's session. Obviously, that's not happening if I'm getting timeouts of 60 minutes or less. Hopefully, the headend can take a look at their settings and get this squared away expeditiously.

I have to take my hat off to the Charter techs (home and line) in my area who have done an outstanding job trying to help track this down. It's ending up to be out of their hands... but they have been consistent and thorough in their work. They seemed to appreciate all the diagnostic help that I was able to do as well. I have nothing but praise for them thus far. This has been a very weird, unique issue that doesn't appear to be too typical.

And a thank you to all those here on DSLR who have thrown out ideas and theories to help guide the way. It is appreciated!

I am hoping the headend stays on this one to correction just as the techs have seen this issue to a diagnosis beyond them. I'll keep you posted.


beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
reply to nunya
said by nunya:


Nothing in your house talks to a node, it's a device doing light conversion.

Actually in newer SDV systems, node groups use a different RPD than the "@ node" resource manager. E.G. the headend doesn't give a rats ass or need to be bothered if another subscriber is added to the stream locally.
It's not simply a repeating device as you allude. It's a separate ARPD, and therefore vulnerable to failure (hardware or software).

Jim - have you talked to any of your neighbors?

You still are not talking to the node, you are talking to a card in the Head End
--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


SwedishRider
Rider on the Storm
Premium
join:2006-01-11
not Sweden
kudos:1
reply to nunya
Headend looked into their end... they don't know what's going on- everything checked out okay.

They want me to replicate, let the box stay frozen overnight, and then they will log into it in the morning and see what's going on. My box just froze at 4:33 PM... and the headend leaves at 4:30.

Bummer... it'll still be there in the morning I suppose..