OpenWeb ADSL Feedback Thread (Pt2)

[aspruyt]

Apparently, Blizzard really is taking quite a knock! They should have us host their servers on our new Cloud network....I think I have said to much!

*hides*

But my Afrihost account to Blizzard servers are purring...

It feels like I am being shaped after hours more than during office hours when the limited edition product which was sold to me states that shaping only occurs during business hours.

I am on a 2mb line so it is not even a possibility that I can download ridiculous amounts, which I dont do in any way.

From tests the line is extremely bursty and has massive jitter with OW but not with Afrihost.

I will see how it turns out as I have been with OW now for a while (since Afrihost moved to MTN network and they dropped the ball heavily for a time), otherwise it is time to move on again as I do not appreciate being sold one thing and getting something very different down the line.

I have sent a ticket to support like cableguy suggested in the first reply.

Edit 1:

The only other thing I can think about is that IS is having issues.

Edit 2:

I got a reply from support that I should do a tracert to openweb.co.za when this happens again. I fail to see how this will help as the issue seems to be international traffic on IS or OW shaping international traffic because of load the more I think about it.

Edit 3:

Received an alternative routing account from support which I will test over the weekend once I get home from work.

I must also say the support response times at OW really are impressive, almost instant!

 

[Anakha56]

Anything amiss under Current/Incomplete Sessions? http://users.isdsl.net/currentsessions.php

I am going to throw a spanner in your works . I decided to view my session on the IS page and lo and behold, I have two sessions, one in Rondebosch 2 and one in New Germany. So with this you would expect me to be one of those complaining about slow speed and being unable to do anything, correct? Well, I am happy to report that everything is working a hundred percent on my end . By looking at this page I would guess it is the IP refresh we are forced to undergo, if memory serves (which it normally does not), are we not meant to be getting new IP addy's every day at midnight? Or was that a Telkom ISP thing only?

 

[PostmanPot]

I am going to throw a spanner in your works . I decided to view my session on the IS page and lo and behold, I have two sessions, one in Rondebosch 2 and one in New Germany. So with this you would expect me to be one of those complaining about slow speed and being unable to do anything, correct? Well, I am happy to report that everything is working a hundred percent on my end .

I'm glad to hear you aren't having speed problems.

But no, that's not what we would expect. The shared connection you see could simply be an ADSL modem connected at someone's home, idling while they're out or at work.

In other words, to see a speed problem, the other person would have to be using their line.

What's interesting to note, is that out of the fraction of OW clients participating in this thread, and an even smaller fraction of clients who would check IS, you are the 3rd/4th person to report having two active sessions in the past 2/3 days.

Would you mind posting screenshots showing the times connected and the ESRs, please?

Please also keep an eye on the IS page and let us know.

 

[DrJohnZoidberg]

Anything amiss under Current/Incomplete Sessions? http://users.isdsl.net/currentsessions.php

This site is definitely not an accurate measurement. I have records from my pfSense firewall and the IS site doesn't reflect anywhere near the correct usage.

IS:

Actual Usage (Feb):

Hence the reason MrBeep says not to use this as a measurement tool.

 

[PostmanPot]

This site is definitely not an accurate measurement.

Hence the reason MrBeep says not to use this as a measurement tool.

We are not measuring usage/anything.

 

[Tinuva]

This site is definitely not an accurate measurement. I have records from my pfSense firewall and the IS site doesn't reflect anywhere near the correct usage.

IS:

Actual Usage (Feb):

Hence the reason MrBeep says not to use this as a measurement tool.

Well yours is the 1st I see where the IS site report less than what you have used. That said, your PFSense value suspiciously look like almost double the value recorded by IS site. Is it possible your pfsense is recording more than 1 interface? eg. multiple adsl accounts, or perhaps the ADSL side as well as the internal interface?

When I had my openweb account, I used Observium's port billing on the specific PPPoE session, and it always recorded less than what the IS side recorded when there were multiple connections on my account.

 

[DrJohnZoidberg]

We are not measuring usage/anything.

Why would you trust it for any data if it's shown not to be accurate?

 

[Draciuz]

I have been with Openweb for a little over a year now.2mbps Gold account.Over the past few weeks i expect to see torrents running at full speed after 9pm but this has not been the case.Yesterday I was getting 10kbps on a torrent, and then i switched to my axxess paid capped account and low and behold it was running at 400kbps.Put in my cancellation yesterday."Uncapped" in this country is a damn sham.

That's exactly why I cancelled with Openweb, and purchased an Afrihost business account. 24/7 I get maximum speeds throughout all protocols. Not exactly the cheapest, but at least you pay for what you get. Afrihost's support is fantastic.

 

[PostmanPot]

You ask I deliver. BTW If I suffer any performance drop I am blaming you because prior to this post everything is awesome!

Thank you. You didn't blank out your username entirely, but I'm not sure how to remove an attachment from a post. I will ask a mod to try.

If you could please repost with all usernames blanked out.

 

[PostmanPot]

Why would you trust it for any data if it's shown not to be accurate?

You will find out soon enough.

And we are in the process of proving that the usage is accurate, at least in some cases (I won't say all yet). Please see Tinuva's post.

*edit*

I would also add that you are one of the few people in the country on 20/40Mbps.

 

[DrJohnZoidberg]

Well yours is the 1st I see where the IS site report less than what you have used. That said, your PFSense value suspiciously look like almost double the value recorded by IS site. Is it possible your pfsense is recording more than 1 interface? eg. multiple adsl accounts, or perhaps the ADSL side as well as the internal interface?

When I had my openweb account, I used Observium's port billing on the specific PPPoE session, and it always recorded less than what the IS side recorded when there were multiple connections on my account.

Nope, this is only for the WAN interface. Vnstat confirms this too:


Note: I only started logging this interface through vnstat halfway through Feb, but you can see the full amount for Mar which is already way over the amount logged in the IS usage stats.

nfe0  /  monthly

       month        rx      |     tx      |    total    |   avg. rate
    ------------------------+-------------+-------------+---------------
      Feb '14    221.59 GiB |   13.09 GiB |  234.67 GiB |  813.74 kbit/s
      Mar '14    333.69 GiB |   20.07 GiB |  353.76 GiB |    2.52 Mbit/s
    ------------------------+-------------+-------------+---------------
    estimated    758.18 GiB |   45.61 GiB |  803.79 GiB |

Here is daily usage to confirm this:

 nfe0  /  daily

         day         rx      |     tx      |    total    |   avg. rate
     ------------------------+-------------+-------------+---------------
      02/13/14     11.36 GiB |  370.97 MiB |   11.72 GiB |    1.14 Mbit/s
      02/14/14     12.32 GiB |  754.37 MiB |   13.06 GiB |    1.27 Mbit/s
      02/15/14     21.54 GiB |  581.08 MiB |   22.10 GiB |    2.15 Mbit/s
      02/16/14     19.92 GiB |  843.91 MiB |   20.74 GiB |    2.01 Mbit/s
      02/17/14      5.52 GiB |  400.01 MiB |    5.91 GiB |  573.47 kbit/s
      02/18/14      4.04 GiB |  484.05 MiB |    4.51 GiB |  438.06 kbit/s
      02/19/14      3.03 GiB |  241.60 MiB |    3.27 GiB |  317.26 kbit/s
      02/20/14      2.21 GiB |  377.28 MiB |    2.58 GiB |  250.27 kbit/s
      02/21/14    875.92 MiB |  554.59 MiB |    1.40 GiB |  135.63 kbit/s
      02/22/14      2.68 GiB |  126.69 MiB |    2.81 GiB |  272.48 kbit/s
      02/23/14      6.57 GiB |  414.08 MiB |    6.97 GiB |  676.74 kbit/s
      02/24/14     11.55 GiB |  743.34 MiB |   12.28 GiB |    1.19 Mbit/s
      02/25/14     12.67 GiB |  690.28 MiB |   13.34 GiB |    1.30 Mbit/s
      02/26/14     14.83 GiB |    1.68 GiB |   16.51 GiB |    1.60 Mbit/s
      02/27/14     16.49 GiB |    1.06 GiB |   17.55 GiB |    1.70 Mbit/s
      02/28/14     16.56 GiB |    1.83 GiB |   18.39 GiB |    1.79 Mbit/s
      03/01/14     26.73 GiB |    1.53 GiB |   28.26 GiB |    2.74 Mbit/s
      03/02/14     17.72 GiB |    1.47 GiB |   19.19 GiB |    1.86 Mbit/s
      03/03/14     12.22 GiB |    1.64 GiB |   13.86 GiB |    1.35 Mbit/s
      03/04/14     13.90 GiB |  631.74 MiB |   14.52 GiB |    1.41 Mbit/s
      03/05/14     15.26 GiB |    2.05 GiB |   17.31 GiB |    1.68 Mbit/s
      03/06/14     14.01 GiB |    1.35 GiB |   15.36 GiB |    1.49 Mbit/s
      03/07/14     26.81 GiB |    2.19 GiB |   29.00 GiB |    2.82 Mbit/s
      03/08/14     31.59 GiB |    1.72 GiB |   33.31 GiB |    3.23 Mbit/s
      03/09/14     87.55 GiB |    2.87 GiB |   90.42 GiB |    8.78 Mbit/s
      03/10/14     31.90 GiB |    1.45 GiB |   33.35 GiB |    3.24 Mbit/s
      03/11/14     31.99 GiB |    1.34 GiB |   33.33 GiB |    3.24 Mbit/s
      03/12/14     17.50 GiB |  682.60 MiB |   18.17 GiB |    1.76 Mbit/s
      03/13/14      4.42 GiB |  825.61 MiB |    5.23 GiB |  507.81 kbit/s
      03/14/14      2.07 GiB |  392.49 MiB |    2.45 GiB |  368.67 kbit/s
     ------------------------+-------------+-------------+---------------
     estimated      3.20 GiB |     607 MiB |    3.79 GiB |

 

[mikeygza]

This site is definitely not an accurate measurement.

I think that that site does not measure anything at all. It merely pulls radius accounting data from a database for the username requested, radius accounting data that they (IS) like any other ISP out there using their own radius for authorization receives from SAIX's radius proxies.

So, are you saying that SAIX's radius accounting packets are sending through the wrong information to ISPs out there when you connect up with DSL?

Michael

 

[Anakha56]

Thank you. You didn't blank out your username entirely, but I'm not sure how to remove an attachment from a post. I will ask a mod to try.

If you could please repost with all usernames blanked out.

lol This is why I should not be doing anything like this after just waking up . How I missed those 2 obvious spots I don't know... post deleted which seems to have trashed the attachments as well.

 

[PostmanPot]

Good stuff.

lol This is why I should not be doing anything like this after just waking up . How I missed those 2 obvious spots I don't know... post deleted which seems to have trashed the attachments as well.

 

[DrJohnZoidberg]

I think that that site does not measure anything at all. It merely pulls radius accounting data from a database for the username requested, radius accounting data that they (IS) like any other ISP out there using their own radius for authorization receives from SAIX's radius proxies.

So, are you saying that SAIX's radius accounting packets are sending through the wrong information to ISPs out there when you connect up with DSL?

Michael

I'm just providing the numbers, I've shown that from two separate measuring tools that the IS numbers don't match my usage at all.

 

[Slamz]

Had a problem with my account... got a new username.. end in @mynetwork.co.za.. is this ok? or should i be worried about my speeds....

 

[DrJohnZoidberg]

Had a problem with my account... got a new username.. end in @mynetwork.co.za.. is this ok? or should i be worried about my speeds....

Lol, use it and find out.

 

[Tinuva]

Nope, this is only for the WAN interface. Vnstat confirms this too:


Note: I only started logging this interface through vnstat halfway through Feb, but you can see the full amount for Mar which is already way over the amount logged in the IS usage stats.

nfe0  /  monthly

       month        rx      |     tx      |    total    |   avg. rate
    ------------------------+-------------+-------------+---------------
      Feb '14    221.59 GiB |   13.09 GiB |  234.67 GiB |  813.74 kbit/s
      Mar '14    333.69 GiB |   20.07 GiB |  353.76 GiB |    2.52 Mbit/s
    ------------------------+-------------+-------------+---------------
    estimated    758.18 GiB |   45.61 GiB |  803.79 GiB |

Here is daily usage to confirm this:

 nfe0  /  daily

         day         rx      |     tx      |    total    |   avg. rate
     ------------------------+-------------+-------------+---------------
      02/13/14     11.36 GiB |  370.97 MiB |   11.72 GiB |    1.14 Mbit/s
      02/14/14     12.32 GiB |  754.37 MiB |   13.06 GiB |    1.27 Mbit/s
      02/15/14     21.54 GiB |  581.08 MiB |   22.10 GiB |    2.15 Mbit/s
      02/16/14     19.92 GiB |  843.91 MiB |   20.74 GiB |    2.01 Mbit/s
      02/17/14      5.52 GiB |  400.01 MiB |    5.91 GiB |  573.47 kbit/s
      02/18/14      4.04 GiB |  484.05 MiB |    4.51 GiB |  438.06 kbit/s
      02/19/14      3.03 GiB |  241.60 MiB |    3.27 GiB |  317.26 kbit/s
      02/20/14      2.21 GiB |  377.28 MiB |    2.58 GiB |  250.27 kbit/s
      02/21/14    875.92 MiB |  554.59 MiB |    1.40 GiB |  135.63 kbit/s
      02/22/14      2.68 GiB |  126.69 MiB |    2.81 GiB |  272.48 kbit/s
      02/23/14      6.57 GiB |  414.08 MiB |    6.97 GiB |  676.74 kbit/s
      02/24/14     11.55 GiB |  743.34 MiB |   12.28 GiB |    1.19 Mbit/s
      02/25/14     12.67 GiB |  690.28 MiB |   13.34 GiB |    1.30 Mbit/s
      02/26/14     14.83 GiB |    1.68 GiB |   16.51 GiB |    1.60 Mbit/s
      02/27/14     16.49 GiB |    1.06 GiB |   17.55 GiB |    1.70 Mbit/s
      02/28/14     16.56 GiB |    1.83 GiB |   18.39 GiB |    1.79 Mbit/s
      03/01/14     26.73 GiB |    1.53 GiB |   28.26 GiB |    2.74 Mbit/s
      03/02/14     17.72 GiB |    1.47 GiB |   19.19 GiB |    1.86 Mbit/s
      03/03/14     12.22 GiB |    1.64 GiB |   13.86 GiB |    1.35 Mbit/s
      03/04/14     13.90 GiB |  631.74 MiB |   14.52 GiB |    1.41 Mbit/s
      03/05/14     15.26 GiB |    2.05 GiB |   17.31 GiB |    1.68 Mbit/s
      03/06/14     14.01 GiB |    1.35 GiB |   15.36 GiB |    1.49 Mbit/s
      03/07/14     26.81 GiB |    2.19 GiB |   29.00 GiB |    2.82 Mbit/s
      03/08/14     31.59 GiB |    1.72 GiB |   33.31 GiB |    3.23 Mbit/s
      03/09/14     87.55 GiB |    2.87 GiB |   90.42 GiB |    8.78 Mbit/s
      03/10/14     31.90 GiB |    1.45 GiB |   33.35 GiB |    3.24 Mbit/s
      03/11/14     31.99 GiB |    1.34 GiB |   33.33 GiB |    3.24 Mbit/s
      03/12/14     17.50 GiB |  682.60 MiB |   18.17 GiB |    1.76 Mbit/s
      03/13/14      4.42 GiB |  825.61 MiB |    5.23 GiB |  507.81 kbit/s
      03/14/14      2.07 GiB |  392.49 MiB |    2.45 GiB |  368.67 kbit/s
     ------------------------+-------------+-------------+---------------
     estimated      3.20 GiB |     607 MiB |    3.79 GiB |

You are looking at the wrong interface on your pfsense.

I did a Google, and nfe0 is the NVIDIA nForce network adapter. For all we know, that adapter is doing both the sharing on your network and the pppoe dialup connections, thus doubling your actual usage.

Thus the data you provide is not valid for proving the IS site data is incorrect. Unless you show us data which is directly coupled to the PPPoE dialup interface. I am going to say, you are just one of the lucky ones that there is only a single person on your account, if it is on the @openweb.co.za realm.

Had a problem with my account... got a new username.. end in @mynetwork.co.za.. is this ok? or should i be worried about my speeds....

That realm is fine, there will only be one user on the account. When I was given a @mynetwork.co.za account, I was the only person ever on it.

 

[mika1000]

back to normal. thanks cableguy

 

[DrJohnZoidberg]

You are looking at the wrong interface on your pfsense.

I did a Google, and nfe0 is the NVIDIA nForce network adapter. For all we know, that adapter is doing both the sharing on your network and the pppoe dialup connections, thus doubling your actual usage.

Thus the data you provide is not valid for proving the IS site data is incorrect. Unless you show us data which is directly coupled to the PPPoE dialup interface. I am going to say, you are just one of the lucky ones that there is only a single person on your account, if it is on the @openweb.co.za realm.

That realm is fine, there will only be one user on the account. When I was given a @mynetwork.co.za account, I was the only person ever on it.

I think I would know my own setup. The only thing connected to this interface is the modem, there is nothing else connected here. I have a pfsense box with two interfaces, this is the WAN interface I've posted for.

Here is the data for the LAN interface, although this is basically the same as this is just a firewall.

re0  /  monthly

       month        rx      |     tx      |    total    |   avg. rate
    ------------------------+-------------+-------------+---------------
      Jan '14     16.12 GiB |  262.14 GiB |  278.26 GiB |  871.50 kbit/s
      Feb '14     21.68 GiB |  415.78 GiB |  437.46 GiB |    1.52 Mbit/s
      Mar '14     19.85 GiB |  332.34 GiB |  352.20 GiB |    2.50 Mbit/s
    ------------------------+-------------+-------------+---------------
    estimated     44.97 GiB |  752.82 GiB |  797.79 GiB |

To reinforce this my SABNZBd usage alone this month is 145GB, that is already over what is reported on IS for this month.