Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[kitz]

>> FEC errors after 12 hours 60,956,219

OUCH
- but at least the line is still hanging in there, and if its only FECs thats high - then you shouldnt be seeing any problems like you would if it was CRCs.

Youve got a pretty damn decent sync speed for the length of your line. - i take it youve done all the usual stuff to improve your line

[weezer]

Thanks to you yes, including removing ring wire, fitting filtered faceplate, shielded router cable and sticking phone on the extension far away from router.
Fitting the filtered faceplate did wonders for my errors but now after 24 hours I have :

Error Seconds (Local/Remote):   1,687 / 0
FEC Errors (Up/Down):   0 / 256,959,131 
CRC Errors (Up/Down):   0 / 5,936
HEC Errors (Up/Down):   0 / 5,000

I had the CRC/HEC errors down to under 1000 for 10 days before I got Tiscali'd.

[kitz]

>> 256,959,131

crazy indeed.  Are you transfering a lot of data and whats your SNRM like now.

[weezer]

SNR margin now down to 5db, lowest it's been for months 

Since last post I now have

FEC Errors (Up/Down):   0 / 315,161,957
CRC Errors (Up/Down):   0 / 23,218
HEC Errors (Up/Down):   0 / 20,553

Son gaming online and has got through 1 GB in 24 hours but that's standard.

Found some old stats from earlier this year:

After 8 days

Connection rate 4160kbps

FEC UP/Down  964 / 3234682
CRC                    0 / 1563
HEC                    0 / 1340

[kitz]

>> SNR margin now down to 5db, lowest it's been for months 

With the amount of errors that youre racking up and the fact your SNRM is far lower than normal... then it looks like your current sync speed is perhaps just a bit too high for your line to be able to cope with in the evenings.

The more data thats transfered.. then obviously the sooner the high errors rack up.

Unless the errors are causing you noticeable slowdowns then you should be ok... but bear in mind that the high error rate will be causing your router to work pretty damn hard.. so dont be too surprised if it feels quite hot.

If it continues to be a problem, you could try doing a reboot in the evening when SNRM is likely to be at its lowest.. it will give you a slower connection speed, but the line should in theory be more stable.

[weezer]

The high numbers were starting to freak me out so did indeed reboot and now have:

Bandwidth (Up/Down) [kbps/kbps]:   448 / 3,744
Data Transferred (Sent/Received) [KB/MB]:   295.00 / 2.62
Output Power (Up/Down) [dBm]:   12.0 / 18.5
Line Attenuation (Up/Down) [dB]:   31.5 / 54.0
SN Margin (Up/Down) [dB]:   16.0 / 8.0

Seems a lot lower connection rate than I've had before with that snr margin, even when rebooting at night and I've been watching these stats for months now (v sad I know).

[kitz]

SNRM fluctuates during the day... normally its at its best in the morning.. and it slowly declines as the evening wears on. 

You could try a reboot at around or just before teatime tomorrow to see if you can get a middle ground.
I wouldnt reboot again today as I dont want the DLM to kick in.  It also records your error rates so with you having millions of FECs you line may be a bit more vulnerable than normal, and may kick in sooner.

[weezer]

Yes, will leave it alone for a while.

Here's another set o figures I had from earlier,

uptime 8 days again
 
connection rate 4544 kbps
FEC               530 / 706574
CRC                  10 / 5195
HEC                   4 / 4871

Just enough for 4 meg profile and margin was never under 6 db (faulty filter sent me back right up to 15 db) using routerstats.

[weezer]

Just a quick update re datastream + FECs. Eclipse recently moved me back to IPstream but before that I took my stats for a typical day -

Uptime:   0 days, 23:56:04
Bandwidth (Up/Down) [kbps/kbps]:   448 / 3,168
Data Transferred (Sent/Received) [MB/MB]:   146.36 / 460.67
Line Attenuation (Up/Down) [dB]:   31.5 / 53.0
SN Margin (Up/Down) [dB]:   16.0 / 15.5
Loss of Framing (Local/Remote):   0 / 0
Loss of Signal (Local/Remote):   0 / 0
Loss of Power (Local/Remote):   0 / 0
Loss of Link (Remote):   0
Error Seconds (Local/Remote):   0 / 0
FEC Errors (Up/Down):   0 / 282,782
CRC Errors (Up/Down):   0 / 27
HEC Errors (Up/Down):   0 / 25

and after -
   
Uptime:   1 day, 9:36:03
Bandwidth (Up/Down) [kbps/kbps]:   448 / 3,104
Data Transferred (Sent/Received) [MB/MB]:   302.67 / 903.90
Line Attenuation (Up/Down) [dB]:   31.5 / 53.0
SN Margin (Up/Down) [dB]:   16.0 / 15.5
Loss of Framing (Local/Remote):   0 / 0
Loss of Signal (Local/Remote):   0 / 0
Loss of Power (Local/Remote):   0 / 0
Loss of Link (Remote):   0
Error Seconds (Local/Remote):   0 / 0
FEC Errors (Up/Down):   0 / 9,238
CRC Errors (Up/Down):   0 / 53
HEC Errors (Up/Down):   0 / 50

Hopefully downstream snr margin will start to come down again fingers crossed.

Edit: Also ppp connection no longer drops every day.

[kitz]

Theoretically speaking the amount of FEC's shouldnt vary depending on which dslam youre on...  although there could be a problem with the linecard on the dslam that youre attached to, but then you would expect it to show other symptoms too.
I am glad that whatever it was though appears to have reduced greatly.

IPStream <-> Datastream could well make a difference re the PPP drops as they will have different routing.