Xilo and Be LLU

Has anyone on Xilo migrated to the Be LLU product? If so, how are things (e.g. latency / speed throughput / etc)?

As a current customer of Xilo (on C&W LLU Pro 16), what's the procedure to move to from the C&W LLU product to the Be LLU (Be 16 or Be 24) product?

Thanks,

Have you asked xilo?

I raised a ticket into Xilo this morning about the charges / migration times but haven't had an answer yet.

I seem to recall a couple of members were discussing it a while back and just wondered if they had moved across and if so, what were their thoughts / experiences.

Edited by aquilla (Fri 16-Dec-11 22:38:41)

Sorry for not coming back to you sooner, the guys in the office were a little hectic today so support queries were prioritised over sales.

If you are with us on Pro 16, you can move to a Be product free of charge and comments from customers on the new service have been very very positive.

Matt

Hi Matt,

Thanks for this. I did drop the query to Sales rather than Support.

How do I go about requesting the migration and what information is needed? You (Xilo) can reply to the open ticket I have if you prefer.

I'll drop you a reply on this shortly before I get to the airport. We can sort this out for you internally though.

Matt

Thanks Matt.

Quick Update.

I migrated to "Be 16" early this morning and so far so good.

I was using a Netgear DG834 and that had synced at 12Mbps following the change over (with a tweaked SNR). I've since moved back to my Cisco 877W (with the latest ADSL patch) and things are looking stable. Maybe try dropping the SNR a bit.

And the connection speed is?

On the Cisco it was just above 9.5Mbps but that was with a default SNR of 9.0dB.

In reply to a post by aquilla:

On the Cisco it was just above 9.5Mbps but that was with a default SNR of 9.0dB.

A bit of a high default Tsnr i thought the norm was 6db?

Not on C&W, their default is 9db (and TTB is 12db).

Matt

Edited by uno (Thu 12-Jan-12 22:35:29)

In reply to a post by tommy45:

In reply to a post by aquilla:

On the Cisco it was just above 9.5Mbps but that was with a default SNR of 9.0dB.

A bit of a high default Tsnr i thought the norm was 6db?

When I connected the Cisco, it reported the SNR was 9.0dB however I understand that on the Be* product, you can email Xilo and they can make target adjustments (possibly all the way down to 3.0dB?).

I also have a DG834G running DGTeam firmware and I can tweak the SNR on that.

If you do tweak SNR down to 3.0dB, the download transfer rate will be dropping fallback. Best to stay on default 9.0dB to get full maximum download transfer rate without dropping fallback.

Yer wo', max?

Firstly that is incomprehensible even by your standards, secondly it is probably total rubbish as all lines are different and many are fine on a 3dB margin.

3dB will get alots of CRC errors on the line!

I have my line set to 3db and have no more errors than with 6, or 9db.

Matt

Down Up
SNR (dB): 3.9 6.1
Attn(dB): 26.0 13.5
Pwr(dBm): 0.0 12.7
Max(Kbps): 21748 1168
Rate (Kbps): 18272 1156

Would you call this a lot?

Total time = 1 days 9 hours 58 min 14 sec
SF = 7522561
CRC = 142
LOS = 0
LOF = 0
ES = 114

I wouldn't!

There is impossible with zero CRC errors on the line ! No ones will get perfect zero no errors on the line. BT Engineerring told me! The best with errors free is set highest snr at between 9 and 12dB.

Mine here:

Noise Margin: 3.3 dB
Connection Rate: 17014 Kbps
Line Attenuation: 32.5 dB
Power: 0.0 dBm
Max Rate: 19932 Kbps

SuperFrames: 37844
SF (CRC) Errors: 37401
Reed Solomon: 0
RS Corrected: 0
RS Un-Corrected: 0
HEC: 26619
Errored Seconds: 2244
Severe ES: 693

Interleave Depth: 1
Bitswaps: 38




ADSL Link Downstream Upstream
Connection Speed 17014 kbps 1279 kbps
Line Attenuation 32.5 db 16.6 db
Noise Margin 3.3 db 7.0 db

Edited by adslmax (Fri 20-Jan-12 00:06:34)

I never said zero, I said no more than a higher margin.... but I get very few even on 6db.

Matt

In reply to a post by adslmax:

.... BT Engineerring told me!

That's called False Authority Syndrome

Ah ok Matt. I think my line is more errors but I just ingorne it. Cos I want the fastest highest speed of 15 Meg. Due to BT profile limit 15000K but why can't BT set 17000K profile to match sync rate ?

Edited by adslmax (Fri 20-Jan-12 00:05:53)

EDIT: Since deleted:

In reply to a post by adslmax:

System Up Time 01:09:31

You have extremely high ES, CRC and HEC for 1h 10m!

Has it re-synced w/out doing a reboot? Some routers reset the Uptime but do not zeroise the error counts at the same time making it difficult to analyse them.

Edited by XRaySpeX (Fri 20-Jan-12 00:14:36)

I just ingorne the errors ! It won't harm my broadband line! It use to be under 25 CRC in an hour before but I think my local exchange had a major uplift shift ! Use to be line attenuation of 34dB before but now reduce to 32.5dB.

Edited by adslmax (Fri 20-Jan-12 00:21:03)

Then it is quite likely that your assertion is true: On a line highly susceptible to errors, like yours, you get more CRC errors at 3 dB NM than at higher NMs.

But that is not generalisable to more normal lines, like mine, as 3 of us have confirmed

In reply to a post by adslmax:

Ah ok Matt. I think my line is more errors but I just ingorne it. Cos I want the fastest highest speed of 15 Meg. Due to BT profile limit 15000K but why can't BT set 17000K profile to match sync rate ?

Your stats look as though nearly every SuperFrame (37401) has at least one uncorrected error. Given the Error Seconds and Severe Error Seconds you had a lot more than 37401 error packets within those SuperFrames.

Each errored packet has to be transmitted again. That means your real data speed goes down.

If you were using Interleaved instead of Fast Path, most of the errors would be corrected and not need retransmitting. Your real data speed would be higher.

I don't like interleaved on because it getting high ping in gaming! Fast path reduced ping!

In reply to a post by adslmax:

I don't like interleaved on because it getting high ping in gaming! Fast path reduced ping!

Logic fail by you!

If you are getting that rate of error packets then a fair number of them will be gaming-related. Each will require retransmission. That knackers your gaming. The actual gaming latency will be highly variable, rather than what a short ping test may show. Have you tried a 50-packet ping test?

All comments I have seen here from avid gamers has been that connection speed is far less important, once you get to these levels, than predictably steady latency. Your connection speed only exists only for a few kilometres to the exchange. Your underlying latency also depends on the hundreds or thousands of mile to the gaming server; the efficiency of every router it goes through; and the performance of the gaming server.

Interleaving at your end would almost certainly be better for your gaming. Possibly even better would be a noise margin of 6dB or 9dB to eliminate the vast majority of errors in the first place, as that way Fast Path may actually be of benefit to you rather than mess you up.

Since my last post, I appear to have found a suitable combination of IOS & Firmware patch for the Cisco 877W that will allow it to operate at ADSL2+ without issues on Xilo's Be 16 package.

The connection has been running fine for about 6 days now. The logs show the connection has dropped 5 times in that period (although 2 of those were me when I moving some stuff around).

Using an IOS command I've tweaked the SNR down to around 3.0dB and it appears to be happily sitting at about 10.5Mbps / 1.1Mbps sync. Speedtests are showing about 9Mbps throughput downstream and 1Mbps upstream.

As for errors, the Cisco is recording errors but I think there might be a bug sorry feature which causes the error count to read incorrectly (e.g. the FEC ES Errors on the Upstream which is an enormous figure the moment I connect, and the Reed-Solomon EC on the upstream).

Q: Can anyone offer any insight into why I am only getting 10Mbps on a line with 34.0dB attenuation? At the moment I'm connected to the upstairs extension but I get the same reading (and same sync speed) in the test socket. Tried different filters and routers but get the same. If that's the way it is then fine.

Text

1 23 45 67 89 1011 1213 1415 1617 1819 2021 2223 2425 2627 2829 3031 3233 3435 3637 3839 4041 4243 4445 4647 4849 5051

rtr-01>show ver | i uptime rtr-01 uptime is 5 days, 20 hours, 58 minutes  rtr-01>show dsl int atm 0ATM0 Alcatel 20190 chipset information                ATU-R (DS)                      ATU-C (US) Modem Status:    Showtime (DMTDSL_SHOWTIME)DSL Mode:        ITU G.992.5 (ADSL2+) Annex A ITU STD NUM:     0x03                            0x2Chip Vendor ID:  'STMI'                          'BDCM' Chip Vendor Specific:  0x0000                    0xA1E9Chip Vendor Country:   0x0F                      0xB5 Modem Vendor ID: 'CSCO'                          'BDCM'Modem Vendor Specific: 0x0000                    0x0000 Modem Vendor Country:  0xB5                      0xB5Serial Number Near:    FCZxxxxxxZ 877W-G-E 15.0 Serial Number Far:  Chip ID:     C196 (0)DFE BOM:         DFE3.0 Annex A (1) Capacity Used:   96%                             99%Noise Margin:     2.0 dB                          5.5 dB Output Power:    19.0 dBm                        12.5 dBmAttenuation:     34.0 dB                         16.0 dB FEC ES Errors:    0                              85287ES Errors:       891                             301 SES Errors:      192                             10LOSES Errors:     0                               1 UES Errors:      13822                            0Defect Status:                       LOM     None Last Fail Code:  NoneWatchdog Counter: 0x05 Watchdog Resets: 0Selftest Result: 0x00 Subfunction:     0x00Interrupts:      226914 (0 spurious) PHY Access Err:  0Activations:     5 LED Status:      ONLED On Time:     100 LED Off Time:    100Init FW:         init_AMR_4.0.223.bin Operation FW:    AMR-E-4.0.223.binFW Source:       external FW Version:      4.0.223                   DS Channel1      DS Channel0   US Channel1       US Channel0Speed (kbps):             0            10615             0              1188 Cells:                    0        674143975             0         239263002Reed-Solomon EC:          0                0             0         595655620 CRC Errors:               0              925             0               678Header Errors:            0              912             0             29293

Can anyone offer any insight into why I am only getting 10Mbps on a line with 34.0dB attenuation?

Similar with the Netgear it seems, so not the router. I'd expect 13Mbps+ on a 6dB margin.

You do mean the test socket inside the master? Not the normal external socket.

Are you using any phone extension cables to reach the sockets, and have you tried different filters?

If you remove the master faceplate, do all extensions stop working?

Have you got a phone-connected burglar alarm?

Just brought myself a brand new Netgear DGND3700 Wireless-N600 Dual Band Gigabit ADSL2+ Wireless Modem Router and I was very impressed with the highest sync rate, thanks to this high performance router.

Before on ADSL:

ADSL Link Downstream Upstream
Connection Speed 8128 kbps 448 kbps
Line Attenuation 30.5 db 16.1 db
Noise Margin 6.2 db 9.3 db

Now on ADSL2+:

ADSL Link Downstream Upstream
Connection Speed 18739 kbps 1279 kbps
Line Attenuation 32.5 db 17.4 db
Noise Margin 3.3 db 6.3 db

http://www.speedtest.net/result/1724403220.png

Hi Robert,

You do mean the test socket inside the master? Not the normal external socket.

Yes, the test socket inside the master socket.

Are you using any phone extension cables to reach the sockets, and have you tried different filters?

There is only one extension (upstairs) and that is wired in properly to the back of the master socket downstairs. I'm then using a standard RJ11 cable from the router to this socket.

If you remove the master faceplate, do all extensions stop working?

Yes.

Have you got a phone-connected burglar alarm?

No. Haven't got Sky either. Just two phones (one per extension) and the broadband. All filtered using ADSLNation filters (no faceplates). Tried replacement filters and other brands but still the same.

Edit: I have also done the quiet line test (17070) from both the test socket and the extension and it's fine - no noise. I have a feeling this is just the way it's going to be.

Edited by aquilla (Mon 23-Jan-12 20:14:54)

In reply to a post by aquilla:

Q: Can anyone offer any insight into why I am only getting 10Mbps on a line with 34.0dB attenuation? At the moment I'm connected to the upstairs extension but I get the same reading (and same sync speed) in the test socket. Tried different filters and routers but get the same. If that's the way it is then fine.

That's the way it is. To get max sync -12Mbps? - you need a Broadcom-based router.

Hi Batboy,

As I thought. So long as the Cisco box remains stable and happy then I'm happy. I would replace the Cisco but I'm a Cisco engineer so I like to use what I work with.

Ask Xilo to put you on the Service-Profile 20 (INP1) which is for none broadcom chipsets.

Edited by deleted (Tue 24-Jan-12 20:36:03)

In reply to a post by HighBeta:

Ask Xilo to put you on the Service-Profile 20 (INP1) which is for none broadcom chipsets.

@HighBeta,

What exactly is "Service-Profile 20" and how will it affect the connection (advantages / disadvantages)?

In reply to a post by HighBeta:

Ask Xilo to put you on the Service-Profile 20 (INP1) which is for none broadcom chipsets.

I'm pretty sure the Cisco supports extended INP No real need to switch it off.

http://www.support.draytek.co.uk/kb_vigor_inp.html

Adding to what BatBoy says, there is an extension to INP that is implemented by Broadcom, which is why a Broadcom-chipped router is best, (apart from one poster's experience), on Be which has Broadcom line cards and defaults to use that extension.

What is being said here is that, (I don't know but it looks logical), the Service Profile 20 does not activate the extension, so routers that don't support that extension may work better than if it is activated.

Thank goodness for Broadcom routers, I just moved in to a new property and signed up for Xilo Be 16. My Netgear DG834GT reports 60.5db downstream attenuation but with a DGTeam firmware and the noise margin tweaked to 50% of default I'm getting a stable 4.5Mb.

Thanks for all the information guys. I may put this to Xilo and see what happens.

Spoke to Xilo last night. INP1 is not available on the Be 16 package that I'm on, it's either INP2 (current setting) or INP0.

With regards to the router, Cisco say this:

Q. Does the Cisco 870 and 850 series support extended INP functions?
A. All the platforms support INP, but only Cisco 877M supports the extended INP functions.

Q. What do I need to know about INP support on the Cisco 870 and 850 series?
A. With Cisco 877 and Cisco 857, customers may experience lower than expected downstream rate when INP is enabled on the DSLAM. This is due to a limitation with the MTK20196 chipset. This issue is addressed with the Cisco 877M platform, using the newer MTK20196P chipset.

In reply to a post by aquilla:

Spoke to Xilo last night. INP1 is not available on the Be 16 package that I'm on, it's either INP2 (current setting) or INP0.

With regards to the router, Cisco say this:

Q. Does the Cisco 870 and 850 series support extended INP functions?
A. All the platforms support INP, but only Cisco 877M supports the extended INP functions.

Q. What do I need to know about INP support on the Cisco 870 and 850 series?
A. With Cisco 877 and Cisco 857, customers may experience lower than expected downstream rate when INP is enabled on the DSLAM. This is due to a limitation with the MTK20196 chipset. This issue is addressed with the Cisco 877M platform, using the newer MTK20196P chipset.

So is your Cisco 877W on the 877M platform, "using the newer MTK20196P chipset?"

In reply to a post by BatBoy:

In reply to a post by aquilla:

Spoke to Xilo last night. INP1 is not available on the Be 16 package that I'm on, it's either INP2 (current setting) or INP0.

With regards to the router, Cisco say this:

Q. Does the Cisco 870 and 850 series support extended INP functions?
A. All the platforms support INP, but only Cisco 877M supports the extended INP functions.

Q. What do I need to know about INP support on the Cisco 870 and 850 series?
A. With Cisco 877 and Cisco 857, customers may experience lower than expected downstream rate when INP is enabled on the DSLAM. This is due to a limitation with the MTK20196 chipset. This issue is addressed with the Cisco 877M platform, using the newer MTK20196P chipset.

So is your Cisco 877W on the 877M platform, "using the newer MTK20196P chipset?"

Nope. My Cisco 877W is an ordinary Annex A router.

Wow i'm jelly

great stats

for the line attenuation.

ADSL Link Downstream Upstream
Connection Speed 20031 kbps 1279 kbps
Line Attenuation 31.5 db 17.4 db
Noise Margin 0.2 db 6.7 db

I was surprise my new Netgear DGND3700 600N can hold very low snr of 0.2dB for 5 days now without loss of sync or disconnection. Got excellent speed of throughput 17 Meg now. http://speedtest.net/result/1736219771.png

Edited by adslmax (Mon 30-Jan-12 13:55:15)