Print

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

[Ronski]

It looks like I started using the new scripts that log more data on the 6/09/2012 at 22:46

I think it's possibly disk activity that's causing the problem, I've looked at the modem_stats log, and the minutes are out of sync at around 14:30 today, and also yesterday at 13:52 - 14:03

Now the server does back up all the other PC's at around 14:00, but the data is written to other drives, but that doesn't mean Windows is not accessing the C drive.

GetStats take 11 seconds to run on my office PC, which has a slightly slower CPU, but an SSD, whilst on the server it takes about 18 seconds, which has a slightly faster CPU, but a normal HDD, possibly even a 5400RPM one.

I added the following lines of code, as the first and last lines of code in  Getstats to measure the time.

echo Start Time %time%>C:\HG612_Modem_Stats\Ongoing_Stats\starttime.txt

echo End Time   %time%>>C:\HG612_Modem_Stats\Ongoing_Stats\starttime.txt

Time for Holby 

[Ronski]

Spoke to my neighbour this morning, from two doors up the road, so a bit further away from the cabinet and he reckons he's getting around 55 meg on speed tests. He said I can unlock his modem sometime, presuming it's an HG612 and graph the stats. Which will be very handy for comparing mine too.

[Bald_Eagle1]

Spoke to my neighbour this morning, from two doors up the road, so a bit further away from the cabinet and he reckons he's getting around 55 meg on speed tests. He said I can unlock his modem sometime, presuming it's an HG612 and graph the stats. Which will be very handy for comparing mine too.

That will indeed be handy, for comparison purposes.
Is he also with Plusnet? If so, that would be a really good comparison.

Incidentally, I noticed the sample from 01:25 this morning is missing & the one from 01:26 took a long time to obtain, hence the slightly more than double "spike" in some of the stats.

I presume that was a harvesting issue rather than a connection issue as error seconds & other error counts didn't show a sudden increase.

Just for curiosity, Does the Task Scheduler history show that it actually ran/started at that time?
On the odd occasion that I see a missing sample in my own logs (or don't see it as it is missing  ), it appears that "something" prevented Task Scheduler from running.
I only ever see this during virus scans & even then it is not every time a scan ruuns on its daily schedule.

[Ronski]

The eagle is very eagle eyed, that was me, I rebooted the server.

For some reason the task schedule that runs the batch file for the daily graphs at midnight didn't run, and also I had two cmd processes running and a strange error code for the scheduled task, 0x41301 I think. So I rebooted and run the task manually, but still got an error code of 0x1, and a cmd process left with the task showing as still running.

Just figured it out, we updated the graphpd batch file, but I forgot to take care of the pause at the end.

As for my neighbour, I'm not sure who he's with, forgot to ask.

[Ronski]

There seems to be some action on the graphs yesterday afternoon/evening.

Dropbox graphs



PS. No problems with the logging that I've noticed, I've even enable teststats to run at midday and midnight (30 seconds past).

[Ronski]

Just as a little update, also keeps this thread as a diary when anything strange happens.

Noticed two resyncs yesterdays, second one seemed to reset the modem, it's uptime reset and so did the time/date.



It seems interleaving very slightly increased each time.

DS D3 power went negative and D3 signal attenuation went from 0 to 64, so could it now be using D3??

There seemed to be a few other changes in yesterdays graphs as well.

Looking at the Bit loading graph, I do seem to have a bit of band D3 kicked in, not much but it's going in the right direction. Speeds have altered but only slightly.

[Bald_Eagle1]

Noticed two resyncs yesterdays, second one seemed to reset the modem, it's uptime reset and so did the time/date.

Also, Retrain Reason is 0, which suggests either a reboot or a power disconnection/reconnection (mometary power cut maybe?)

Retrain Reason is usually 2 for quick "on the fly" re-syncs that don't cause a modem reboot or trigger a new PPP session.

[Bald_Eagle1]

Noticed two resyncs yesterdays, second one seemed to reset the modem, it's uptime reset and so did the time/date.

Also, Retrain Reason is 0, which suggests either a reboot or a power disconnection/reconnection (mometary power cut maybe?)

Retrain Reason is usually 2 for those quick, "on the fly" re-syncs that don't cause a modem reboot or trigger a new PPP session.

[Bald_Eagle1]

Noticed two resyncs yesterdays, second one seemed to reset the modem, it's uptime reset and so did the time/date.

Also, Retrain Reason is 0, which suggests either a reboot or a power disconnection/reconnection (mometary power cut maybe?)

Retrain Reason is usually 2 for those quick, "on the fly" re-syncs that don't cause a modem reboot or trigger a new PPP session.

[Ronski]

No power cuts, was watching TV a recording off the server at the time, also no problem with the UPS - just checked the server logs and all was fine there. So only place a power problem could of occured was the modem or it's power supply, which is unlikely.

My FTTC weirdness continues......

[Bald_Eagle1]

My FTTC weirdness continues......

Hmmm,

I wonder if Plusnet or even BT could have caused a reboot via the TR 069 back door - for whatever reason.

Even reboots usually take less than one minute hence no disruption/missing stats from the modem_stats.log either.

 

[Ronski]

Had a resync this morning at 10am, speed has increased again, D3 antenuation has dropped to 63.1 from 64, and more bits being used.

If keeps on like this I might get back to a proper speed, eventually.........

The attainable rate doesn't seemed to have increased much, but at least the sync is now at 45 Meg.

[Bald_Eagle1]

I wonder if the colder weather has contributed at all.

It has been rather frosty for a couple of days where I live & my sync speed & attainable rates have increased slightly via an "on the fly" resync at 01:03 this morning.

[Ronski]

Yes, it'd be interesting to see if the reverse happens in the spring, someone said (possibly you) that it might be the minute contractions of the copper wires affecting the joints, making slightly better connections.

[Bald_Eagle1]

I do believe I may well have said that.
I do make it all up as I go along though  .

I have also read that radio propagation a.k.a. "skip" interference is less during colder (therefore usually autumn or winter) months. 

Many years ago I recall occasionally conversing with people in France & Spain on a basic & very low powered FM CB radio during summer months, but never during winter.

My own experiences have shown that VDSL2 "problems" occurred more often during warm & dry weather than in cold & wet weather.