User comments on ISPs
  >> EE (Everything Everywhere) and Orange


Register (or login) on our website and you will not see this ad.


Pages in this thread: 1 | [2] | 3 | >> (show all)   Print Thread
Standard User XRaySpeX
(eat-sleep-adslguide) Fri 03-Apr-15 23:53:08
Print Post

Re: I thought it was a major outage but maybe not?


[re: Kail] [link to this post]
 
In reply to a post by Kail:
Those stats are after a reboot which means it has added the current stats to what went before
Quite the opposite! A physical reboot (switch OFF/ON) resets all error counts and Time Connected to 0. That's why I asked my Q. Your 2nd set of stats 2 mins after a reboot proves this. Even those ES are high but it's too short a period to say.

However your 3rd & 4th sets of stats prove that there is a helluva lotta errors on your line all the time.

As a rule of thumb on Interleaved I would expect no more than 1 ES every 3 mins. You are getting 1 ES in just over 1 sec. consistently.

1999: Freeserve 48K Dial-Up => 2005: Wanadoo 1 Meg BB => 2007: Orange 2 Meg BB => 2008: Orange 8 Meg LLU => 2010: Orange 16 Meg LLU => 2011: Orange 20 Meg WBC
Standard User XRaySpeX
(eat-sleep-adslguide) Sat 04-Apr-15 00:03:03
Print Post

Re: I thought it was a major outage but maybe not?


[re: flippery] [link to this post]
 
In reply to a post by flippery:
The low SNR down will not help with stability or errors. Increasing to 6 may.
The BTw DLM will run with a default target NM = 3 dB quite happily on good stable lines. The Q is why it is seeing the OP's as such?

Yes, throughput can be influenced by the ISP's network, but their fibre networks are highly unlikely to damp speeds down to a fraction of a Meg.

1999: Freeserve 48K Dial-Up => 2005: Wanadoo 1 Meg BB => 2007: Orange 2 Meg BB => 2008: Orange 8 Meg LLU => 2010: Orange 16 Meg LLU => 2011: Orange 20 Meg WBC
Standard User David_W
(fountain of knowledge) Sat 04-Apr-15 02:24:09
Print Post

Re: I thought it was a major outage but maybe not?


[re: XRaySpeX] [link to this post]
 
As XRaySpeX has pointed out, you have 42297 error seconds in 13:53:19 uptime (49999 seconds), which appears to indicate just under 85% of seconds are near end errored seconds.

Edit: The remainder of this post is mistaken and I withdraw it. Please see the correction posted below.

However, there are two cautions in interpreting this number.


Firstly, routers are inconsistent over which direction is near end and far end when it comes to reporting errors. The DSL standards indicate that far end statistics relate to the errors detected in what the equipment is receiving and near end are the errors reported by the remote end in what the equipment are sending. This means near end statistics reported by customer equipment should relate to the upstream and far end to the downstream, but some routers report things the opposite way round, from the perspective of the DSLAM.

If the router is following the DSL standards, the problem is in the upstream direction.


Secondly, some DSLAMs don't reset the counts they send following resynchronisation (which would be the near end statistics if the DSL standard definitions are followed). However, this problem does not apply here, as the data shows the number of error seconds clearly goes back to 0 following resynchronisation.


It is impossible to tweak the upstream noise margin even if you have tweakable hardware. You can only tweak the downstream noise margin. This is a shame, because it would be interesting to push the upstream noise margin higher, to see if the proportion of seconds that are errored seconds decreases.


The question is what can be done. You've tried different routers and filters. It is possible the problem is noise getting into your extension cabling or directly into your router and then on to the line. If you have any powerline adaptors, unplug them and see if the proportion of errored seconds goes down, though this is a bit of a long shot (I'd expect problems with powerline adaptors to affect the downstream more than the upstream).

I don't have the practical experience of others in these forums in troubleshooting network side issues. However, remembering that the upstream signal is weakest at the DSLAM end, my suspicion is that the problem may be a defective port on the DSLAM or excessive noise being picked up at or near the DSLAM (possibly due to crosstalk). A 'lift and shift' might well help.


Edited by David_W (Sat 04-Apr-15 10:48:46)


Register (or login) on our website and you will not see this ad.

Standard User XRaySpeX
(eat-sleep-adslguide) Sat 04-Apr-15 03:28:50
Print Post

Re: I thought it was a major outage but maybe not?


[re: David_W] [link to this post]
 
In reply to a post by David_W:
the data shows the number of error seconds clearly goes back to 0 following resynchronisation.
Not resync but reboot! Only Time Connected is 0'ed upon resync.

The Brightbox does report downstream as 'Near End'. Otherwise I wouldn't be discussing it. However please clarify:
near end are the errors reported by the remote end in what the equipment are sending.
Which equipment? The remote-end equipment (DSLAM) about itself? Or the customer equipment (modem)? If the latter, the remote end will be reporting on what itself is receiving in order for it to be upstream comms. I find this sentence confusing as to equipment & direction.

1999: Freeserve 48K Dial-Up => 2005: Wanadoo 1 Meg BB => 2007: Orange 2 Meg BB => 2008: Orange 8 Meg LLU => 2010: Orange 16 Meg LLU => 2011: Orange 20 Meg WBC
Standard User David_W
(fountain of knowledge) Sat 04-Apr-15 04:45:18
Print Post

Re: I thought it was a major outage but maybe not?


[re: XRaySpeX] [link to this post]
 
In reply to a post by XRaySpeX:
However please clarify:
near end are the errors reported by the remote end in what the equipment are sending.
Which equipment? The remote-end equipment (DSLAM) about itself? Or the customer equipment (modem)? If the latter, the remote end will be reporting on what itself is receiving in order for it to be upstream comms. I find this sentence confusing as to equipment & direction.


There's no ambiguity in "The DSL standards indicate that far end statistics relate to the errors detected in what the equipment is receiving and near end are the errors reported by the remote end in what the equipment are sending." Unfortunately, it's backwards (it's late, sorry)!

G.992.1, the ADSL standard (which is a starting point for the later standards) says:

3.18 far-end: Far-end means performance of the downstream loop-side received signal at the input of the ATU-R, where this performance is reported to the ATU-C in upstream indicators (see Figure 9-4), or performance of the upstream loop-side received signal at the input of the ATU-C, where this performance is reported to the ATU-R in downstream overhead indicators; this case is a mirror image of the above (see Figure 9-4).

3.23 near-end: Near-end means performance of the loop-side received signal at the input of the ATU (see Figure 9-4).


Figure 9-4 is not that helpful, but the meaning of the definitions is clear. Near end relates to what the equipment receives. From the perspective of ATU-R, the subscriber side, near end is the downstream.

Far end relates to what is reported to the equipment by the other side about what the other side is receiving. From the perspective of ATU-R, the subscriber side, far end is what ATU-C, the DSLAM, reports about the signal it is receiving - in other words, from the perspective of ATU-R, far end is the upstream.


The original poster has a near end errored seconds problem, which is a downstream problem assuming these definitions are followed. Please ignore my earlier mistaken suggestion about this being an upstream problem.

Standard User XRaySpeX
(eat-sleep-adslguide) Sat 04-Apr-15 05:30:04
Print Post

Re: I thought it was a major outage but maybe not?


[re: David_W] [link to this post]
 
Ah, that's more like it. Thanks! It'll be as well to delete or edit your preceeding post so as not to confuse OP or subsequent readers.

1999: Freeserve 48K Dial-Up => 2005: Wanadoo 1 Meg BB => 2007: Orange 2 Meg BB => 2008: Orange 8 Meg LLU => 2010: Orange 16 Meg LLU => 2011: Orange 20 Meg WBC
Standard User Kail
(member) Sat 04-Apr-15 18:53:00
Print Post

Re: I thought it was a major outage but maybe not?


[re: XRaySpeX] [link to this post]
 
In reply to a post by XRaySpeX:
t'll be as well to delete or edit your preceeding post so as not to confuse OP or subsequent readers.
Overwhelmed head spinning, etc. wink

In reply to a post by David_W:
Near end relates to what the equipment receives.

So it comes down to which way round the Bright Box displays the information and if EE follow the standard? Thanks, think I've got that?

So if I'm right... The Bright Box displays Upstream then Downstream. The next section display Near End then Far End. So looking at the screen it looks like the right column is Downstream and Far End together when Downstream is actually (or should be if they are doing it right) the left column marked Near End meaning what my Bright Box is receiving?

I've had another re-sync and more errored seconds while the internet was not in use this afternoon, about 16:00 while we were all out, performance appears to be currently normal...

Internet
Broadband Type ADSL
Broadband State Connected
Internet State Connected
Time Connected 02:46:46
Downstream Rate 5.44 Mbps
Upstream Rate 1.04 Mbps
Status
Configured Current
Line Status -- UP
Link Type -- Interleaved Path
Operation Mode Automatic G.992.5 (ADSL2+)
Data Rate Information
Upstream 1044 (Kbps)
Downstream 5440 (Kbps)
Defect/Failure Indication
Operation Data Upstream Downstream
Noise Margin 6.2 (dB) 6.0 (dB)
Line Attenuation 24.9 (dB) 47.5 (dB)
Indicator Name Near End Indicator Far End Indicator
Output Power 12.8 (dBm) 0.0 (dBm)
Fast Path FEC Correction -- --
Interleaved Path FEC Correction 4299373 244
Fast Path CRC Error -- --
Interleaved Path CRC Error 0 71
Loss of Signal Defect 1 0
Fast Path HEC Error STR -- --
Interleaved Path HEC Error 100 26
Error Seconds 121002 354
Statistics
Received Data 658984 (Kbits)
Transmitted Data 51749 (Kbits)
Standard User XRaySpeX
(eat-sleep-adslguide) Sat 04-Apr-15 19:09:18
Print Post

Re: I thought it was a major outage but maybe not?


[re: Kail] [link to this post]
 
In reply to a post by Kail:
Downstream is actually (or should be if they are doing it right) the left column marked Near End meaning what my Bright Box is receiving?
Yes, Downstream is 'Downstream' in the top part and 'Near End' in the bottom part.

All you need to know is that you have an extraordinarily high downstream ES (& other error counts but ES is sufficient as a summation) all the time.
while the internet was not in use this afternoon
FYI: Even when you are not using the Net, packets are still passing constantly at the sync rate; they just contain no user data, but may still error.

1999: Freeserve 48K Dial-Up => 2005: Wanadoo 1 Meg BB => 2007: Orange 2 Meg BB => 2008: Orange 8 Meg LLU => 2010: Orange 16 Meg LLU => 2011: Orange 20 Meg WBC

Edited by XRaySpeX (Sat 04-Apr-15 19:14:44)

Standard User XRaySpeX
(eat-sleep-adslguide) Sun 05-Apr-15 00:24:42
Print Post

Re: I thought it was a major outage but maybe not?


[re: Kail] [link to this post]
 
Did you ever tell us your exchange?

1999: Freeserve 48K Dial-Up => 2005: Wanadoo 1 Meg BB => 2007: Orange 2 Meg BB => 2008: Orange 8 Meg LLU => 2010: Orange 16 Meg LLU => 2011: Orange 20 Meg WBC
Standard User eckiedoo
(experienced) Sun 05-Apr-15 07:32:26
Print Post

Re: I thought it was a major outage but maybe not?


[re: flippery] [link to this post]
 
Thinking further about it, the replacement 8-way cable may simply be a better quality.
Pages in this thread: 1 | [2] | 3 | >> (show all)   Print Thread

Jump to