dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
6543
ameridian
join:2013-06-23
Paterson, NJ

ameridian

Member

MODEM (Arris TM822G) rebooting practically everyday

I have OV and based on the MTA section of the event log it appears as though the MODEM is rebooting at least once per day. The relevant log entries do coincide with reboots if I'm around to check things when they occur. Have Ultra 100/35. Anyone else notice this?
frdrizzt
join:2008-05-03
Ronkonkoma, NY

frdrizzt

Member

How do your signals look? Do you see any uncorrectables or correctables incrementing in the status page? Any other logs (T3/T4 timeouts)? Any video issues too?
TheWiseGuy
Dog And Butterfly
MVM
join:2002-07-04
East Stroudsburg, PA

1 edit

TheWiseGuy to ameridian

MVM

to ameridian
There are many causes of spontaneous reboots but the most likely is a signal related issue. Even if your signals look good it could still be intermittent noise. The log entries might give a clue if it is a signal issue.

Robin Walkers reference page offer some of the possibilities if it is not a signal related issue.

»homepage.ntlworld.com/ro ··· l#reboot
ameridian
join:2013-06-23
Paterson, NJ

1 edit

ameridian to frdrizzt

Member

to frdrizzt
said by frdrizzt:

How do your signals look? Do you see any uncorrectables or correctables incrementing in the status page? Any other logs (T3/T4 timeouts)? Any video issues too?

Thanks for responding, here is the status page and relevant event logs;

»ic2.pbase.com/o9/96/4259 ··· atus.png
»ic2.pbase.com/o9/96/4259 ··· csis.png
»ic2.pbase.com/o9/96/4259 ··· mca1.png
»ic2.pbase.com/o9/96/4259 ··· mca2.png

The only tangible issues are no internet or VOIP access during the period the MODEM decides to reboot. Event ID sequence 3, 65517....3, 14 in the MTA logs represent a reboot sequence.
ameridian

ameridian to TheWiseGuy

Member

to TheWiseGuy
Thanks for the link. The Upstream power levels have always been on the high side but I can't recall seeing it above 54.00 dBmV. I called Optimum on 8/15 and one of the Techs took a look at things then rebooted the MODEM.

He didn't see anything out of the ordinary and said their utility there indicated that the MODEM had been up continuously for 17 days but confessed that their tool doesn't always track uptime correctly for some reason. He also told me to see how things fared after he rebooted it. Well, as you can see from the last log link in my previous post, the MODEM rebooted itself less than 24 hours after he rebooted it on 8/15 (6:19 PM).
TheWiseGuy
Dog And Butterfly
MVM
join:2002-07-04
East Stroudsburg, PA

TheWiseGuy

MVM

The T-3 and T-4 errors are likely due to a signal issue. Since the main symptom seems to be reboot, it is likely an intermittent issue with something occurring at the time of the reboot. You could try posting in the direct forum, they may be able to look at the flap list to see if there are other modems rebooting at the same time.

»Optimum Direct

RARPSL
join:1999-12-08
Suffern, NY

RARPSL to ameridian

Member

to ameridian

Up time Display
said by ameridian:

He didn't see anything out of the ordinary and said their utility there indicated that the MODEM had been up continuously for 17 days but confessed that their tool doesn't always track uptime correctly for some reason.

If the modem is rebooting once a day, the claim that it has been up for 17 days is odd since as shown in this capture the time since the last boot is shown and the tech should be able to see it. In your case it should be under 1 day (go to »192.168.100.1 to see this display).
ameridian
join:2013-06-23
Paterson, NJ

ameridian to TheWiseGuy

Member

to TheWiseGuy
said by TheWiseGuy:

The T-3 and T-4 errors are likely due to a signal issue. Since the main symptom seems to be reboot, it is likely an intermittent issue with something occurring at the time of the reboot. You could try posting in the direct forum, they may be able to look at the flap list to see if there are other modems rebooting at the same time.

»Optimum Direct

After further investigation the problem seems to be VOIP related. I had suspected this for sometime but didn't have the time to dig deeper until today. In other words the reboots seem to be triggered shortly after VOIP use, after hanging up the phone, but it doesn't happen all the time. I was able to cause two reboots within 20 minutes by fiddling with the phones today;

»ic2.pbase.com/o9/96/4259 ··· mca3.png

The house is wired for two phone lines that are disconnected from Verizon's external setup. We have two VOIP lines connected to 3 dual line phones and one single line phone. All the phones are cordless setups. The Arris VOIP line is connected to one line of the 3 dual line phones and a Magic Jack + is connected to the other line. The MJ+ is also connected to the single line phone. Don't know if the wiring is causing noise or other issues for the Arris but the previous Motorola MODEM had no issues. Furthermore, the MJ+ is covering a wider area, is used much more often and that has no issues. Going to do some more fiddling tomorrow.
ameridian

ameridian to RARPSL

Member

to RARPSL
said by RARPSL:

If the modem is rebooting once a day, the claim that it has been up for 17 days is odd since as shown in this capture the time since the last boot is shown and the tech should be able to see it. In your case it should be under 1 day (go to »192.168.100.1 to see this display).

Odd indeed, but all indications are that they don't see the same status page and logs we see. Spoke with a Tech today and setup a service visit for Tuesday. I asked him some questions during the conversation and he basically indicated that he was using some other tool for diagnosis so we were never on the "same page".

RARPSL
join:1999-12-08
Suffern, NY

RARPSL

Member

said by ameridian:

said by RARPSL:

If the modem is rebooting once a day, the claim that it has been up for 17 days is odd since as shown in this capture the time since the last boot is shown and the tech should be able to see it. In your case it should be under 1 day (go to »192.168.100.1 to see this display).

Odd indeed, but all indications are that they don't see the same status page and logs we see. Spoke with a Tech today and setup a service visit for Tuesday. I asked him some questions during the conversation and he basically indicated that he was using some other tool for diagnosis so we were never on the "same page".

I would not know. OTOH, if they are interrogating the modem, they SHOULD have access to the same information (including the field used to populate that display section).
ameridian
join:2013-06-23
Paterson, NJ

ameridian

Member

A Tech came today unannounced, it was my wife who alerted me to his presence. He was taking some measurements where the underground cable comes up and is coupled to the cable going inside the house. His diagnosis was that the old RG-59U cable had to be replaced due to poor signal levels reaching the modem.

We went through this around 5 years ago when I first got internet from Optimum. Not only did they declare that the cable had to be replaced back then, but they also had someone come and paint up my driveway along with the road in the townhouse complex in preparation to dig. Luckily, I was working from home when they arrived one day with a large truck and auger looking contraption on it. Turns out the problem was where the street cable connects to the underground cable on the pole, the connection box was saturated with water.

The fellow this time around lied through his teeth, insisting that he had determined that loss/attenuation/degradation was taking place only within the underground cable run (around 100'). So I asked him how he could determine this without going up the pole. Also, read him the riot act, telling him there is no way in hell Optimum is going to be digging up my driveway or the complex.

Long story short, about 30 minutes after the altercation, he goes up on the pole and services the connection box and these are the before/after results;

Before:
»m3.i.pbase.com/o9/96/425 ··· atus.png

After:
»ic2.pbase.com/o9/96/4259 ··· tus1.png

Hopefully, this has solved the reboot problem. If not, it's probably a flaky modem.
ameridian

1 edit

ameridian

Member

Two cable TV boxes and modem meant using 2 splitters and the modem was off the second splitter. Optimum had located it, around 5 years ago, where the previous DSL modem was. This was based on my suggestion since it would be just a matter of plug and play in terms of network and VOIP connections. But what would signals be like if there was no splitter or the modem was off the first splitter?

Took some status screen shots;

No splitter:
»ic2.pbase.com/o9/96/4259 ··· tus2.png

Off one splitter:
»ic2.pbase.com/o9/96/4259 ··· tus3.png

The results were so promising that I relocated modem and router downstairs so the modem was off the first splitter instead of the second. Network and VOIP rerouting changes were minimal. Here is the final result with everything rewired;

»ic2.pbase.com/o9/96/4259 ··· tus4.png

So the tap box servicing and modem relocation has resulted in a 13 dBmV increase in downstream power levels and an 11 dBmV decrease in upstream power levels. Those Arris status screens are a godsend to the consumer.
cablewizzard
join:2009-06-14
Woodbury, NY

2 edits

cablewizzard

Member

You were losing 7.5 dB between the output of the first splitter, plus cable run to the 2nd splitter, and through the 2nd splitter?

That's quite a bit. Or your 2nd splitter was a 3-way with -7/-7/-3.5 , and your modem happened to be wrongly plugged into one of the two -7 dB ports: 3.5 dB wasted.

Yes, the tap fixing got you some 5 dB back on downstream, and seemingly even more (6-7 dB) on the return - but a full half of your signal loss was way sub-optimal wiring in your house, which should have been addressed on the very first truck roll to your house (which I understand didn't happen, they tried to address the outdoor issues first, which is NOT standard procedure).

So yes, the status screens are useful (but field techs are not trained in them, the vast majority of them does NOT have laptops - the smarter ones tend to at least KNOW about them).

Technical service reps have their own tools that are going beyond what the WWW status screens can show. It still doesn't mean they understand or interpret the results correctly: a +54dB transmit is an UPPER LIMIT that the modem must not and will not exceed (if bonded with 2 US channels; the single-channel/unbonded limit in DOCSIS3.0 is +57 for the US channel type CV is using (6.4Mhz/64QAM).

Generally, reps are not trained to recognize that a +54 dBmV transmit is undesirable, as that may still not be powerful enough to reach the head-end with +-0dBmV - indeed the signal at the headend is likely BELOW 0, but the modem dutifully reports +54dBmV, and that's considered "within acceptable range".

Likewise for field techs with their test equipment: that equipment WILL tell them Rx/Tx power for any channel they desire (be that a video or DOCSIS QAM channel) - but likewise: if the Tx power is +54dBmV, it still means the head-end side can be below 0dBmV, but there's no way for a device in the field to find out about that (the CMTS does not tell the modem the signal strength it receives from it, it only asks the modems to make adjustments based on its own calculations).

I have to admit that the first tech on the phone should have recognized the -13dBmV downstream Rx power as out of CV spec (+10 to -10, despite DOCSIS permitting +15 to -15), it should have been named an RF/Signal issue at that point, and caused a truck roll dispatched to your house without any further need for arguments - you having to call back again (and set up the truck roll appointment) was undesirable from a customer experience point of view.

Last: the 17 days of uptime on your modem (that the first rep unduly over-emphasized) are the modem's system uptime , e.g.: SNMP RFC1213-MIB::sysUpTime.0 , identical to what is on the WWW status page. Just because the DOCSIS interface in the modem was forced to re-initialize (lost signal long enough for 16 T3 errors to pile up -> T4 -> reinit) does not mean the modem did a full reset all the way down to the boatloader (which would reset the system uptime). There's situations, like the modem going into "partial service" (losing the bonding) and then coming back out of it works without a full system reset: Yet, you experience it as a reboot (because the modem re-entered initial ranging, turning all but the power-light off). In some vendor's equipment, this behavior can be explicitly configured.
ameridian
join:2013-06-23
Paterson, NJ

ameridian

Member

Using two 2-way splitters. Yes, 10 dB drop for both and 7 dB for just one seems odd since these are normally 3.5 dB attenuation per leg. Both are Optimum supplied.

I've found the DOCSIS events log for this modem to be next to useless in determining when a spontaneous reboot had occurred. OTOH, the PacketCable events log clearly indicates this. Also, as you pointed out, spontaneous reboots do not reset the uptime counter. Probably they shouldn't be called reboots at all if by definition a reboot is supposed to reset the uptime counter. A case of semantics I guess.

andrewc2
join:2011-06-05
Matamoras, PA

andrewc2 to ameridian

Member

to ameridian
I saw your other post in Boost's thread, and I can't believe that so many people are against using the conduit... To me conduit is far superior to direct burial of cable...