|
|
|
|
| Author |
Message |
Timothy Daniels
Guest
|
 Posted: Thu Dec 28, 2006 8:37 am Post subject: Re: SB5120 & Comcast Woes |
 |
|
"Timothy Daniels" wrote:
| Quote: | "Todd H." wrote:
"Timothy Daniels" writes:
I'm afraid that I'd have to show that the speeds are dependent on
the downstream signal level, and right now that doesn't seem to be
the case. The techs seemed to be happy when they dropped it from
15 dBmV to 9 - 10 dBmV (measured at the modem with a barrel
connector between the modem and the wall).
This barrel connector... is it a standard female to female coupler or
does it have any attenuation built in?
No, it's just a standard coupler (it's threaded on both ends on the
outside).
In my experience with this crap, packet loss is what it boils down to.
If you can demonstrate packet loss, and you have eliminated your
router from the situation, they should be sending a tech. Unless you
own your modem, in which case, I'd throw some money at trying a new
modem.
I'll drop by Time Warner tomorrow and try to borrow a modem -
they're only 5 blocks away. If I can eliminate the modem as a
cause of the packet loss, I'll be able to prove it's a network
problem. Otherwise, I'll have to sit on the phone for an hour again
to get Tier 3 to order another tech visit and the guy may or may not
show up. Line techs have been here, but they don't do anything
about network problems, and the tech with the substitute modem
10 days ago didn't try any pings or tracerts.
*TimDaniels*
|
OK, to make a looooong story short, I've substituted a Terayon
modem from Time Warner in place of my Motorola SB5120,
and surprise - no timeouts! With more than 200 pings sent to
various servers, not one lost packet. With my 5120, I was getting
sometimes up to 20% lost packets, usually somewhere between
5% and 10% - which RoadRunner Tech Help said was normal
and to be expected. <sheeesh>
The speedtest at Speakeasy seems to have crept up a little bit
(from 5.2 to 5.6 Mbps), but the speed at DSL Reports seems
to have gone *down*. But the speeds reported vaied so randomly
that I really don't know if the speeds were affected at all. Now I
wonder if there was really something wrong with the Motorola or
whether the parameters downloaded by RoadRunner were wrong
or just a bit off. I was under the impression that the Motorola
Surfboard 5120 was top of the line, and now I'm disappointed to
think that I may have gotten a lemon.
*TimDaniels* |
|
| Back to top |
|
 |
Fix your Windows Problems - FAST.
FREE Safe Scan Registry Check. Locate & Fix Errors in Minutes!
|
|
Warren
Guest
|
| Posted: Thu Dec 28, 2006 10:25 am Post subject: Re: SB5120 & Comcast Woes |
|
|
Timothy Daniels wrote:
| Quote: | OK, to make a looooong story short, I've substituted a Terayon
modem from Time Warner in place of my Motorola SB5120,
and surprise - no timeouts! With more than 200 pings sent to
various servers, not one lost packet. With my 5120, I was getting
sometimes up to 20% lost packets, usually somewhere between
5% and 10% - which RoadRunner Tech Help said was normal
and to be expected. <sheeesh
The speedtest at Speakeasy seems to have crept up a little bit
(from 5.2 to 5.6 Mbps), but the speed at DSL Reports seems
to have gone *down*. But the speeds reported vaied so randomly
that I really don't know if the speeds were affected at all. Now I
wonder if there was really something wrong with the Motorola or
whether the parameters downloaded by RoadRunner were wrong
or just a bit off. I was under the impression that the Motorola
Surfboard 5120 was top of the line, and now I'm disappointed to
think that I may have gotten a lemon.
|
The new modem has likely caused you to be put on a different node or
lasergroup. So was the problem the modem, or the CMTS? You aren't really
in a position to tell.
Based on everything that has passed, I would say that you had multiple,
unrelated problem -- at least two different problem. As for cause and
effect, well, there have been just too many variables to do more than
speculate.
--
Warren H.
==========
Disclaimer: My views reflect those of myself, and not my
employer, my friends, nor (as she often tells me) my wife.
Any resemblance to the views of anybody living or dead is
coincidental. No animals were hurt in the writing of this
response -- unless you count my dog who desperately wants
to go outside now.
Shop for networking gear:
http://www.holzemville.com/mall/linksys
http://www.holzemville.com/mall/netgear |
|
| Back to top |
|
|
Todd H.
Guest
|
| Posted: Thu Dec 28, 2006 11:17 am Post subject: Re: SB5120 & Comcast Woes |
|
|
"Timothy Daniels" <TDaniels@NoSpamDot.com> writes:
| Quote: | OK, to make a looooong story short, I've substituted a Terayon
modem from Time Warner in place of my Motorola SB5120,
and surprise - no timeouts! With more than 200 pings sent to
various servers, not one lost packet. With my 5120, I was getting
sometimes up to 20% lost packets, usually somewhere between
5% and 10% - which RoadRunner Tech Help said was normal
and to be expected. <sheeesh
|
Heh... Yikes. I certainly wouldn't be very happy with a provider
that only successfully delivered 90-95% of my packets on the first
try.
As for the speed numbers, I wouldn't obsess about them or the
differences because as you said, if you have fantastic download speeds
according to one speed test, but are consistently seeing 20% packet
loss at certain times, it's like having a car that has 400HP engine,
but only will pull away from the traffic light 80% of the time.
| Quote: | I was under the impression that the Motorola Surfboard 5120 was
top of the line, and now I'm disappointed to think that I may
have gotten a lemon.
|
How long did you have it? How's the weather been since you got it?
It may have soaked up one too many power surges and started
failing..you never know. Or as another poster correctly suggested
it's possible that your new modem being activated perhaps masked the
problem via other channels. it's hard to say.
Glad you're in happyville though with the new modem. These
intermittent packet loss problems are a pain in the *** to get
resolved and I'm glad you got where you wanted to go.
Best Regards,
--
Todd H.
http://www.toddh.net/ |
|
| Back to top |
|
|
Ed Nielsen
Guest
|
| Posted: Thu Dec 28, 2006 2:04 pm Post subject: Re: SB5120 & Comcast Woes |
|
|
Congratulations on discovering the timeout issue.
While it is true that, in many people's opinions, the Motorola
surfboard is top of the line, there can be a lemon found anywhere. It
is also possible (I think more likely) that the config file somehow
became corrupted and when the new cable modem was installed and
provisioned, the file was completely rebuilt and the Terayon got the
correct config file.
Speedtests through DSL Reports are always slow for me as compared to
speedtests directly to the same server.
CIAO!
Ed N.
Timothy Daniels wrote:
| Quote: |
OK, to make a looooong story short, I've substituted a Terayon
modem from Time Warner in place of my Motorola SB5120,
and surprise - no timeouts! With more than 200 pings sent to
various servers, not one lost packet. With my 5120, I was getting
sometimes up to 20% lost packets, usually somewhere between
5% and 10% - which RoadRunner Tech Help said was normal
and to be expected. <sheeesh
The speedtest at Speakeasy seems to have crept up a little bit
(from 5.2 to 5.6 Mbps), but the speed at DSL Reports seems
to have gone *down*. But the speeds reported vaied so randomly
that I really don't know if the speeds were affected at all. Now I
wonder if there was really something wrong with the Motorola or
whether the parameters downloaded by RoadRunner were wrong
or just a bit off. I was under the impression that the Motorola
Surfboard 5120 was top of the line, and now I'm disappointed to
think that I may have gotten a lemon.
*TimDaniels* |
|
|
| Back to top |
|
|
Timothy Daniels
Guest
|
| Posted: Thu Dec 28, 2006 2:04 pm Post subject: Re: SB5120 & Comcast Woes |
|
|
"Timothy Daniels" wrote:
| Quote: | "Warren" wrote:
Timothy Daniels wrote:
"Warren" wrote:
Most -- nearly every -- cable provider won't burn a public IP address
on the external interface of a cablemodem since it will never need to
be accessed from outside the company's network. It will be assigned a
private range IP, usually a class A address (10.x.x.x).
Please tell me more about this 10.x.y.z. IP address. That address,
the first upstream node in my tracerts, is the only address that has no
ping dropouts. Is that the modem as seen by the PC? Everything
else upstream has a 2% and higher dropout rate.
You won't see either of the cablemodem's interfaces IP addresses in any
traceroute as the cablemodem is just a bridge. The 10.x.x.x IP address you
are likely seeing is just another of the many aliases of the CMTS. (Or it
could be some other router still within the cable company's network.)
I've again asked the RoadRunner Tier I Nat'l Help Desk what that
IP address represents, and after checking with someone else, the
rep said that it was the modem. (Not that Tier I actually knows
something, but it would explain why pings to it were always returned.)
*TimDaniels*
|
I asked again, and now the rep says (after consultation) that the
10.x.y.Z IP address belongs to "the thing which puts you on the Internet,
the thing near the servers, not in the neighborhood". So it remains a
mystery why my "lossy" Motorola SB5120 wouldn't drop any packets
at all when pinging that 10.x.y.z thing and would drop lots of packets
when pinging nodes upstream of it.
*TimDaniels* |
|
| Back to top |
|
|
|
|
Warren
Guest
|
| Posted: Thu Dec 28, 2006 10:49 pm Post subject: Re: SB5120 & Comcast Woes |
|
|
Timothy Daniels wrote:
| Quote: | I asked again, and now the rep says (after consultation) that the
10.x.y.Z IP address belongs to "the thing which puts you on the
Internet,
the thing near the servers, not in the neighborhood". So it remains a
mystery why my "lossy" Motorola SB5120 wouldn't drop any packets
at all when pinging that 10.x.y.z thing and would drop lots of packets
when pinging nodes upstream of it.
|
Well, connect the dots.
Your signal-level issues and your packet loss issues were completely
unrelated, and your modem was not a problem at all.
The 10.x.x.x address could not have been your modem because your modem is
just a bridge, and the IP address of a bridge will not show in a
traceroute. It is, as I said before, one of the addresses of one of the
interfaces on the CMTS. It is your gateway. Your network configuration (or
your router's configuration if you're using a router) likely shows a
different IP address as your gateway, but it would be the same device.
Despite any signal-level issues you might have been having, you weren't
having any packet loss to the gateway. In other words, even though your
modem had to yell too much to communicate, the messages were all getting
through. As noted earlier in this discussion, a signal-level issue could
have an end result of packet loss, but there is no direct coloration
between signal-level and packet loss.
The problem causing your issues has not been solved. It only no longer
affects you. There is still an interface on a CMTS that is not functioning
properly out there. You're just no longer connected to it because of your
modem swap. Or maybe, by some coincidence, someone fixed the problem
already, too. We don't know. We don't know what network engineering was
looking at or doing at the same time your modem swap was being done. The
assumption that your modem swap was the only thing happening at that point
in time is fundamentally faulty. There could have been a lot more going on
behind the scenes at the same time.
So when it comes down to the bottom line, all you can do is speculate as
to what was happening. Enough variables have been changed to resolve your
issue. You only know about some of the variables that have changed. You
found a number of problems along the way, but there is no logical reason
to make an assumption that any of those problems were related. Trying to
mold some explanation that ties these problems together is an exercise in
futility.
You're also expecting customer service to give you far more information
than they are trained to understand. And it's all irrelevant, too. Does it
really matter what that 10.x.x.x address belongs to? The customer service
agent's job is to determine if a problem is something on your end that can
be resolved over the phone. And if it's not, is it a problem that one of
their techs need to be dispatched for, or something unrelated to the
service. That's it. They don't need to know what that 10.x.x.x address
belongs to. A tech in the field should have some concept of what it is,
but only in the sense that it may have something to do with a problem that
they need to pass on to someone else to look at. So if you're relying on
customer service to give you technical information, you're not
understanding their purpose. And now that the issue has been resolved,
you're just wasting their time drilling them for irrelevant information.
You had an issue. During troubleshooting, multiple problems were found and
fixed. Those multiple problems were not related to each other. Other
variables may have changed due to actions that you have no knowledge of.
Those actions may have been taken as a result of troubleshooting your
issue, or those actions may have been taken for some unrelated issue. The
bottom line is that you no longer have an issue.
There is no need for a huge conference to uncover just what combination of
events were most relevant in this one particular case. This alignment of
the stars will never happen again. Next time there will be some other
alignment of the stars that may result in similar issues. Solving those
theoretical problems in the future will not be facilitated by knowing what
happened this time. That theoretical future problem will best be resolved
by methodical troubleshooting, and fixing each problem found during that
troubleshooting. How your mystery problem was solved this time is
irrelevant to any future issues.
And that's also why it's so infuriating to hear, "I had the same problem,
and I solved it by this obscure action." Each issue has different
variables. Each issue should be methodically troubleshooted. Applying
random "solutions" that worked for someone else who had an issue with
similar symptoms is a waste of time. And thus trying to do a full
inquisition into the minutia of this problem that no longer exists is just
a waste of everyone's time. The problem is resolved, and future problems
won't be resolved faster by abandoning methodical troubleshooting of the
unique situation.
--
Warren H.
==========
Disclaimer: My views reflect those of myself, and not my
employer, my friends, nor (as she often tells me) my wife.
Any resemblance to the views of anybody living or dead is
coincidental. No animals were hurt in the writing of this
response -- unless you count my dog who desperately wants
to go outside now.
Shop for networking gear:
http://www.holzemville.com/mall/linksys
http://www.holzemville.com/mall/netgear |
|
| Back to top |
|
|
Guest
|
| Posted: Fri Dec 29, 2006 1:56 am Post subject: Re: SB5120 & Comcast Woes |
|
|
On Thu, 28 Dec 2006 13:09:09 -0800, "Timothy Daniels"
<TDaniels@NoSpamDot.com> wrote:
| Quote: | What nags at me, Ed, is that the pings with the Motorola
to the CMTS were without any lost packets at all, but that
pings to nodes upstream were losing packets. That
doesn't sound like a modem problem to me, but Road-
Runner Help Desk is so understaffed due to the holidays,
that they require a couple hours in 2 different queues even
at *midnight* Pacific Time to reach someone who can
provision a modem. I guess I'll have to wait until next week
to check this out further.
*TimDaniels*
|
You're probably right when you say they're seriously understaffed, and
you're probably right when you said in a previous post that your
problems seem to be fixed, so why would you consider wasting their
time trying to determine a cause?
As Warren said, there are too many variables involved, some known and
others unknown, to be able to pin down a cause. Just enjoy your new
service and let the dust settle on all of those promises you made
about leaving for DSL. :)
--
Bill |
|
| Back to top |
|
|
Todd H.
Guest
|
| Posted: Fri Dec 29, 2006 1:56 am Post subject: Re: SB5120 & Comcast Woes |
|
|
"Timothy Daniels" <TDaniels@NoSpamDot.com> writes:
| Quote: | What nags at me, Ed, is that the pings with the Motorola
to the CMTS were without any lost packets at all, but that
pings to nodes upstream were losing packets. That
doesn't sound like a modem problem to me
|
I concur. Your data now doesn't suggest a modem issue.
But your problem is fixed. The rest is an academic exercise, which
in my experience with the lowest common denominator types you usually
get to talk to at cable companies makes an academic exercise an
extremely fruitless effort.
--
Todd H.
http://www.toddh.net/ |
|
| Back to top |
|
|
Timothy Daniels
Guest
|
| Posted: Fri Dec 29, 2006 1:56 am Post subject: Re: SB5120 & Comcast Woes |
|
|
"Warren" wrote:
| Quote: | Timothy Daniels wrote:
I asked again, and now the rep says (after consultation) that the
10.x.y.Z IP address belongs to "the thing which puts you on the
Internet,
the thing near the servers, not in the neighborhood". So it remains a
mystery why my "lossy" Motorola SB5120 wouldn't drop any packets
at all when pinging that 10.x.y.z thing and would drop lots of packets
when pinging nodes upstream of it.
Well, connect the dots.
Your signal-level issues and your packet loss issues were completely
unrelated, and your modem was not a problem at all.
The 10.x.x.x address could not have been your modem because your modem is
just a bridge, and the IP address of a bridge will not show in a
traceroute. It is, as I said before, one of the addresses of one of the
interfaces on the CMTS. It is your gateway. Your network configuration (or
your router's configuration if you're using a router) likely shows a
different IP address as your gateway, but it would be the same device.
Despite any signal-level issues you might have been having, you weren't
having any packet loss to the gateway. In other words, even though your
modem had to yell too much to communicate, the messages were all getting
through. As noted earlier in this discussion, a signal-level issue could
have an end result of packet loss, but there is no direct coloration
between signal-level and packet loss.
The problem causing your issues has not been solved. It only no longer
affects you. There is still an interface on a CMTS that is not functioning
properly out there. You're just no longer connected to it because of your
modem swap. Or maybe, by some coincidence, someone fixed the problem
already, too. We don't know. We don't know what network engineering was
looking at or doing at the same time your modem swap was being done. The
assumption that your modem swap was the only thing happening at that point
in time is fundamentally faulty. There could have been a lot more going on
behind the scenes at the same time.
So when it comes down to the bottom line, all you can do is speculate as
to what was happening. Enough variables have been changed to resolve your
issue. You only know about some of the variables that have changed. You
found a number of problems along the way, but there is no logical reason
to make an assumption that any of those problems were related. Trying to
mold some explanation that ties these problems together is an exercise in
futility.
You're also expecting customer service to give you far more information
than they are trained to understand. And it's all irrelevant, too. Does it
really matter what that 10.x.x.x address belongs to? The customer service
agent's job is to determine if a problem is something on your end that can
be resolved over the phone. And if it's not, is it a problem that one of
their techs need to be dispatched for, or something unrelated to the
service. That's it. They don't need to know what that 10.x.x.x address
belongs to. A tech in the field should have some concept of what it is,
but only in the sense that it may have something to do with a problem that
they need to pass on to someone else to look at. So if you're relying on
customer service to give you technical information, you're not
understanding their purpose. And now that the issue has been resolved,
you're just wasting their time drilling them for irrelevant information.
You had an issue. During troubleshooting, multiple problems were found and
fixed. Those multiple problems were not related to each other. Other
variables may have changed due to actions that you have no knowledge of.
Those actions may have been taken as a result of troubleshooting your
issue, or those actions may have been taken for some unrelated issue. The
bottom line is that you no longer have an issue.
There is no need for a huge conference to uncover just what combination of
events were most relevant in this one particular case. This alignment of
the stars will never happen again. Next time there will be some other
alignment of the stars that may result in similar issues. Solving those
theoretical problems in the future will not be facilitated by knowing what
happened this time. That theoretical future problem will best be resolved
by methodical troubleshooting, and fixing each problem found during that
troubleshooting. How your mystery problem was solved this time is
irrelevant to any future issues.
And that's also why it's so infuriating to hear, "I had the same problem,
and I solved it by this obscure action." Each issue has different
variables. Each issue should be methodically troubleshooted. Applying
random "solutions" that worked for someone else who had an issue with
similar symptoms is a waste of time. And thus trying to do a full
inquisition into the minutia of this problem that no longer exists is just
a waste of everyone's time. The problem is resolved, and future problems
won't be resolved faster by abandoning methodical troubleshooting of the
unique situation.
|
Care to summarize that? :-)
*TimDaniels* |
|
| Back to top |
|
|
Timothy Daniels
Guest
|
| Posted: Fri Dec 29, 2006 1:56 am Post subject: Re: SB5120 & Comcast Woes |
|
|
What nags at me, Ed, is that the pings with the Motorola
to the CMTS were without any lost packets at all, but that
pings to nodes upstream were losing packets. That
doesn't sound like a modem problem to me, but Road-
Runner Help Desk is so understaffed due to the holidays,
that they require a couple hours in 2 different queues even
at *midnight* Pacific Time to reach someone who can
provision a modem. I guess I'll have to wait until next week
to check this out further.
*TimDaniels*
"Ed Nielsen" wrote:
| Quote: | Congratulations on discovering the timeout issue.
While it is true that, in many people's opinions, the Motorola
surfboard is top of the line, there can be a lemon found anywhere. It
is also possible (I think more likely) that the config file somehow
became corrupted and when the new cable modem was installed and
provisioned, the file was completely rebuilt and the Terayon got the
correct config file.
Speedtests through DSL Reports are always slow for me as compared to
speedtests directly to the same server.
CIAO!
Ed N.
Timothy Daniels wrote:
OK, to make a looooong story short, I've substituted a Terayon
modem from Time Warner in place of my Motorola SB5120,
and surprise - no timeouts! With more than 200 pings sent to
various servers, not one lost packet. With my 5120, I was getting
sometimes up to 20% lost packets, usually somewhere between
5% and 10% - which RoadRunner Tech Help said was normal
and to be expected. <sheeesh
The speedtest at Speakeasy seems to have crept up a little bit
(from 5.2 to 5.6 Mbps), but the speed at DSL Reports seems
to have gone *down*. But the speeds reported vaied so randomly
that I really don't know if the speeds were affected at all. Now I
wonder if there was really something wrong with the Motorola or
whether the parameters downloaded by RoadRunner were wrong
or just a bit off. I was under the impression that the Motorola
Surfboard 5120 was top of the line, and now I'm disappointed to
think that I may have gotten a lemon.
*TimDaniels* |
|
|
| Back to top |
|
|
Fix your Windows Problems - FAST.
FREE Safe Scan Registry Check. Locate & Fix Errors in Minutes!
|
|
Warren
Guest
|
| Posted: Fri Dec 29, 2006 12:34 pm Post subject: Re: SB5120 & Comcast Woes |
|
|
Timothy Daniels wrote:
| Quote: |
Care to summarize that? :-)
|
The problem is solved. Let it be. There's no need for an M&M conference --
especially since the patient lived, but if you want one, it's not going to
be brief, nor will it be conclusive.
--
Warren H.
==========
Disclaimer: My views reflect those of myself, and not my
employer, my friends, nor (as she often tells me) my wife.
Any resemblance to the views of anybody living or dead is
coincidental. No animals were hurt in the writing of this
response -- unless you count my dog who desperately wants
to go outside now.
Shop for networking gear:
http://www.holzemville.com/mall/linksys
http://www.holzemville.com/mall/netgear |
|
| Back to top |
|
|
|
|
| |
Watched Topics
Search
Register
Log in to check your private messages
Profile
Log in
Memberlist
Usergroups