-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: What's the most interesting small town you've visited in South Africa?
Connection keeps dropping. Help needed.
- Thread starter mackape
- Start date
porchrat
Honorary Master
No way man...I am going to phone them x2 per day and send them 3 e-mails per day until they do something about this...
I think you underestimate their resolve.
mackape
Active Member
I gave them a call and explained the situation. I asked to clear my cookies/history on the browser and when I asked how this would impact on the speed of the download manager, I was refered to on of the techies. Thebogo asked me to flush the DNS using the DOS cmd prompt (this was a bit over my head), but since I was on the phone, I had to tell him I would phone back. I did what I was told, ran a speedtest and the results was up 30KB/s....it is only an average of 72KB/s....but hey, it is better that 40KB/s....I will monitor and phone them again if the problem persists...
Could you guys try the same and give feedback? Or, is there someone who can explain in simple terms why the DNS needed to be flushed to get increased speeds?
Could you guys try the same and give feedback? Or, is there someone who can explain in simple terms why the DNS needed to be flushed to get increased speeds?
Results from Neotel Speedtest site:
Before flush:
Last Result:
Download Speed: 354 kbps (44.3 KB/sec transfer rate)
Upload Speed: 61 kbps (7.6 KB/sec transfer rate)
After Flush:
Download Speed: 850 kbps (106.3 KB/sec transfer rate)
Upload Speed: 64 kbps (8 KB/sec transfer rate)
Download Speed: 686 kbps (85.8 KB/sec transfer rate)
Upload Speed: 65 kbps (8.1 KB/sec transfer rate)
Download Speed: 768 kbps (96 KB/sec transfer rate)
Upload Speed: 59 kbps (7.4 KB/sec transfer rate)
however I'm not able to sustain that speed as downloads from downloadmanager and Grabit are still around 40-50Kb on average.
Before flush:
Last Result:
Download Speed: 354 kbps (44.3 KB/sec transfer rate)
Upload Speed: 61 kbps (7.6 KB/sec transfer rate)
After Flush:
Download Speed: 850 kbps (106.3 KB/sec transfer rate)
Upload Speed: 64 kbps (8 KB/sec transfer rate)
Download Speed: 686 kbps (85.8 KB/sec transfer rate)
Upload Speed: 65 kbps (8.1 KB/sec transfer rate)
Download Speed: 768 kbps (96 KB/sec transfer rate)
Upload Speed: 59 kbps (7.4 KB/sec transfer rate)
however I'm not able to sustain that speed as downloads from downloadmanager and Grabit are still around 40-50Kb on average.
mackape
Active Member
In FDM, change your network options (Tools - Download Options - Network) to the following:
Max # of connections = 12
Max # of connections per server = 12
Max # of running downloads = 2
I have found that this is more reliable than having 20 connections per server and 4 running downloads. Also try and download one file at a time and see the results.
It seems as if the connection is much more stable on the XP machine than the Vista machine. I have uninstalled the NeoManager on the Vista machine so I will give it a try tonight. This trail and error is however very frustrating...when is NeoFlex comming
Max # of connections = 12
Max # of connections per server = 12
Max # of running downloads = 2
I have found that this is more reliable than having 20 connections per server and 4 running downloads. Also try and download one file at a time and see the results.
It seems as if the connection is much more stable on the XP machine than the Vista machine. I have uninstalled the NeoManager on the Vista machine so I will give it a try tonight. This trail and error is however very frustrating...when is NeoFlex comming
Something has changed in the last 4 days. I now have connection times of 62 minutes exactly. My log shows disconnections at 00:23, 01:25, 02:27, 03:29, 4:31, and so it has gone for 4 days now. I haven't touched my setup, and as I've said before, these are forceful disconnections coming from Neotel.
I've faxed them a notice of cancellation and suddenly, after two months of silence, I'm getting offers of assistance. No actual assistance yet - just offers. They want to make me pay for the device, but since no-one has checked to see if the problem is with the device itself I am refusing to buy a potentially defective device.
If they charge me for a device they provided for a service that never worked I'm going to take them to court. And while I'm at it I'll throw in the over-billing issue as well. I've had enough of being stuffed around.
I've faxed them a notice of cancellation and suddenly, after two months of silence, I'm getting offers of assistance. No actual assistance yet - just offers. They want to make me pay for the device, but since no-one has checked to see if the problem is with the device itself I am refusing to buy a potentially defective device.
If they charge me for a device they provided for a service that never worked I'm going to take them to court. And while I'm at it I'll throw in the over-billing issue as well. I've had enough of being stuffed around.
something intressting i noticed
I can be online for 3 to 4 hours .. (playing wow) with no problems
the moment i start a video streaming session i get disconnected
I stop the streaming and the connection is stable again
I can be online for 3 to 4 hours .. (playing wow) with no problems
the moment i start a video streaming session i get disconnected
I stop the streaming and the connection is stable again
I can stream with no disconnections but only get disconnected when playing dota lol, how weird is that. Wanna swop device's ?
mackape
Active Member
My connection has become more "stable" if I can call it that...I have noticed that over the weekend, I have less problems than in the week... I must however admit that I have been checking on the downloads every hour or so (over the weekend). I have stopped downloading during the day and I now only download in the evening and over weekend.
Monad, I seriously share your pain. Are there no lawyers willing to offer their services for a percentage of the settlement fee? I am willing to give 50% of the settlement fee if we can nail Neotel for contract breach (not providing service in terms of contract). In the US or Europe, this would not have happend, and if it did happen, they would throw in a year's worth of free connection...
if I can call it that...I have noticed that over the weekend, I have less problems than in the week... I must however admit that I have been checking on the downloads every hour or so (over the weekend). I have stopped downloading during the day and I now only download in the evening and over weekend.Monad, I seriously share your pain. Are there no lawyers willing to offer their services for a percentage of the settlement fee? I am willing to give 50% of the settlement fee if we can nail Neotel for contract breach (not providing service in terms of contract). In the US or Europe, this would not have happend, and if it did happen, they would throw in a year's worth of free connection...
Last edited:
I finally had an engineer come by to come and have a look. After dropping connections all day long I suddenly had a burst of stability that lasted throughout his visit - he was here for more than an hour. Go figure.
mackape, I agree. In the US service like this would be grounds for a class-action lawsuit. Here people just shrug off bad service as another routine day in Africa. I wonder how much of it is cultural? Do we in Africa have so little respect for others that we don't expect them to respect us?
mackape, I agree. In the US service like this would be grounds for a class-action lawsuit. Here people just shrug off bad service as another routine day in Africa. I wonder how much of it is cultural? Do we in Africa have so little respect for others that we don't expect them to respect us?
You can just ask for one to come over however from my experience all they really do is verify that something is amiss in your area/tower and feed this back to the Engineers - then the engineers conduct their test on the towers when there is hardly any one connected - they claim that all is fine, send you their speedtest results and meanwhile sweet nothing has improved for you.
And this is from experience.
And this is from experience.
How difficult is it to get a technician to come over and take a look?
Im getting timeouts to giganews when I ping through cmd prompt. Is that wierd. This all seems to have started after the massive storm we had here in Bedfordview yesterday. Could that have anything to do with it?
Tho I doubt it cause when i start newsleecher, it heads stright up to 170K for a few seconds and then just drops off back towards zero...its puzzling and driving me mad. Any ideas more advanced, experienced formites?
Im getting timeouts to giganews when I ping through cmd prompt. Is that wierd. This all seems to have started after the massive storm we had here in Bedfordview yesterday. Could that have anything to do with it?
Tho I doubt it cause when i start newsleecher, it heads stright up to 170K for a few seconds and then just drops off back towards zero...its puzzling and driving me mad. Any ideas more advanced, experienced formites?
Just to add something valuable. I just got off the phone from support and spoke a very helpful guy who told me something that makes alot of sense. He said that each of us has a "Profile" that Neotel builds arround our connection. It has to do with the ports we use. If we are recorded as overusing a particular site etc, the port is limited for some reason. So he said that in the begining all sites are treated as guest visits until it reaches a certain traffic level to that site, afterwards you are regarded as a memeber of that site ( so he explained it - no flames please lol ). The solution he said was that he would request all my incoming ports to be opened tomorrow and that should fix any slowdows or problems I had with the sites i was attempting to use, aka Giganews servers. I have been getting over 1500k with the speedtests so its not the connection. All I can say is that the traffic profiling makes a lil more sense to me as to why I cant reach the news servers. The again I have pulled about 20 gig this week already.
Hope that helps someone. Will find out tomoz when the ports are all opened if he was talking out of his tree or not.
Style
Hope that helps someone. Will find out tomoz when the ports are all opened if he was talking out of his tree or not.
Style
Did a reverse traceroute via the Giganews site. Wont lie...I need a bit of help reading it. No that way inclined. Can anyone tell me what probs they see from this info?
1 216.196.98.6 (216.196.98.6) 0 ms 0 ms 0 ms
2 ge-0-2-0-0.was11.ip.tiscali.net (213.200.66.89) 0 ms 0 ms 0 ms
3 xe-1-0-0.was12.ip.tiscali.net (89.149.185.5) 0 ms xe-0-0-0.was12.ip.tiscali.net (89.149.185.1) 0 ms xe-5-0-0.was12.ip.tiscali.net (89.149.185.17) 0 ms
4 213.200.84.82 (213.200.84.82) 9 ms 13 ms 213.200.84.74 (213.200.84.74) 0 ms
5 Vlan40.icore1.NJY-Newark.as6453.net (206.82.132.21) 6 ms 10 ms 17 ms
6 if-1-0-0-1101.mcore4.NJY-Newark.as6453.net (216.6.63.5) 6 ms if-6-0-0-38.mcore4.NJY-Newark.as6453.net (216.6.63.45) 6 ms if-1-0-0-1101.mcore4.NJY-Newark.as6453.net (216.6.63.5) 6 ms
7 if-1-0.mcore3.L78-London.as6453.net (195.219.144.9) 78 ms 80 ms 78 ms
MPLS Label=1794 CoS=0 TTL=0 S=1
8 if-7-0.core2.LDN-London.as6453.net (195.219.144.22) 77 ms 77 ms 83 ms
MPLS Label=238 CoS=0 TTL=0 S=1
9 if-9-0-1.bb2.JSO-Johannesburg.as6453.net (216.6.59.33) 248 ms 247 ms 247 ms
MPLS Label=16 CoS=0 TTL=0 S=1
10 if-1-7.bb1.JSO-Johannesburg.as6453.net (216.6.55.9) 248 ms 247 ms 269 ms
11 ix-1-8.bb1.JSO-Johannesburg.as6453.net (216.6.55.42) 249 ms 247 ms 247 ms
12 41.160.0.243 (41.160.0.2) 249 ms 249 ms 250 ms
13 * * *
14 * * *
15 * * *
Max number of unresponsive hops reached (firewall or filter?)
news-europe.giganews.com
1 vl201.gw3.ams.giganews.com (216.196.110.3) 1 ms 0 ms 0 ms
2 xe-1-2-0-0.ams10.ip.tiscali.net (213.200.75.41) 0 ms 0 ms 0 ms
3 xe-2-2-0.ams20.ip.tiscali.net (89.149.186.202) 0 ms 0 ms 0 ms
4 if-2-1.icore1.AD1-Amsterdam.as6453.net (195.219.150.53) 0 ms 0 ms 0 ms
5 if-7-0-0-1118.core1.AD1-Amsterdam.as6453.net (80.231.80.85) 0 ms if-1-0-0-83.core1.AD1-Amsterdam.as6453.net (80.231.80.5) 0 ms if-7-0-0-1118.core1.AD1-Amsterdam.as6453.net (80.231.80.85) 0 ms
6 if-10-0.mcore3.LDN-London.as6453.net (80.231.80.30) 11 ms 11 ms 13 ms
MPLS Label=452 CoS=0 TTL=0 S=1
7 if-10-0.core2.LDN-London.as6453.net (195.219.195.6) 11 ms 11 ms 12 ms
MPLS Label=238 CoS=0 TTL=0 S=1
8 if-9-0-1.bb2.JSO-Johannesburg.as6453.net (216.6.59.33) 192 ms 183 ms 184 ms
MPLS Label=16 CoS=0 TTL=0 S=1
9 if-1-7.bb1.JSO-Johannesburg.as6453.net (216.6.55.9) 182 ms 196 ms 182 ms
10 ix-1-8.bb1.JSO-Johannesburg.as6453.net (216.6.55.42) 182 ms 182 ms 182 ms
11 41.160.0.243 (41.160.0.2) 183 ms 182 ms 194 ms
12 * * *
13 * * *
14 * * *
Max number of unresponsive hops reached (firewall or filter?)
1 216.196.98.6 (216.196.98.6) 0 ms 0 ms 0 ms
2 ge-0-2-0-0.was11.ip.tiscali.net (213.200.66.89) 0 ms 0 ms 0 ms
3 xe-1-0-0.was12.ip.tiscali.net (89.149.185.5) 0 ms xe-0-0-0.was12.ip.tiscali.net (89.149.185.1) 0 ms xe-5-0-0.was12.ip.tiscali.net (89.149.185.17) 0 ms
4 213.200.84.82 (213.200.84.82) 9 ms 13 ms 213.200.84.74 (213.200.84.74) 0 ms
5 Vlan40.icore1.NJY-Newark.as6453.net (206.82.132.21) 6 ms 10 ms 17 ms
6 if-1-0-0-1101.mcore4.NJY-Newark.as6453.net (216.6.63.5) 6 ms if-6-0-0-38.mcore4.NJY-Newark.as6453.net (216.6.63.45) 6 ms if-1-0-0-1101.mcore4.NJY-Newark.as6453.net (216.6.63.5) 6 ms
7 if-1-0.mcore3.L78-London.as6453.net (195.219.144.9) 78 ms 80 ms 78 ms
MPLS Label=1794 CoS=0 TTL=0 S=1
8 if-7-0.core2.LDN-London.as6453.net (195.219.144.22) 77 ms 77 ms 83 ms
MPLS Label=238 CoS=0 TTL=0 S=1
9 if-9-0-1.bb2.JSO-Johannesburg.as6453.net (216.6.59.33) 248 ms 247 ms 247 ms
MPLS Label=16 CoS=0 TTL=0 S=1
10 if-1-7.bb1.JSO-Johannesburg.as6453.net (216.6.55.9) 248 ms 247 ms 269 ms
11 ix-1-8.bb1.JSO-Johannesburg.as6453.net (216.6.55.42) 249 ms 247 ms 247 ms
12 41.160.0.243 (41.160.0.2) 249 ms 249 ms 250 ms
13 * * *
14 * * *
15 * * *
Max number of unresponsive hops reached (firewall or filter?)
news-europe.giganews.com
1 vl201.gw3.ams.giganews.com (216.196.110.3) 1 ms 0 ms 0 ms
2 xe-1-2-0-0.ams10.ip.tiscali.net (213.200.75.41) 0 ms 0 ms 0 ms
3 xe-2-2-0.ams20.ip.tiscali.net (89.149.186.202) 0 ms 0 ms 0 ms
4 if-2-1.icore1.AD1-Amsterdam.as6453.net (195.219.150.53) 0 ms 0 ms 0 ms
5 if-7-0-0-1118.core1.AD1-Amsterdam.as6453.net (80.231.80.85) 0 ms if-1-0-0-83.core1.AD1-Amsterdam.as6453.net (80.231.80.5) 0 ms if-7-0-0-1118.core1.AD1-Amsterdam.as6453.net (80.231.80.85) 0 ms
6 if-10-0.mcore3.LDN-London.as6453.net (80.231.80.30) 11 ms 11 ms 13 ms
MPLS Label=452 CoS=0 TTL=0 S=1
7 if-10-0.core2.LDN-London.as6453.net (195.219.195.6) 11 ms 11 ms 12 ms
MPLS Label=238 CoS=0 TTL=0 S=1
8 if-9-0-1.bb2.JSO-Johannesburg.as6453.net (216.6.59.33) 192 ms 183 ms 184 ms
MPLS Label=16 CoS=0 TTL=0 S=1
9 if-1-7.bb1.JSO-Johannesburg.as6453.net (216.6.55.9) 182 ms 196 ms 182 ms
10 ix-1-8.bb1.JSO-Johannesburg.as6453.net (216.6.55.42) 182 ms 182 ms 182 ms
11 41.160.0.243 (41.160.0.2) 183 ms 182 ms 194 ms
12 * * *
13 * * *
14 * * *
Max number of unresponsive hops reached (firewall or filter?)
mackape
Active Member
Style, what you are saying is a bit concerning (your thread you posted after speaking to the Neotel consultant). I personally have no problem if Neotel is monitoring internet usage and internet traffic. What I do have a problem with, is when I am throttled if I consistently download from a specific site. Surely that is some sort of contract breach? "We give you uncapped internet, but if you visit a site too often, we are going to cap you".... I am not amused...
Have you tried to "flush" the dns (refer to my previous post whereby you go to the command prompt and type ipconfig /flushdns)? I have noticed a significant increase in speed and I have also had less drops in connection. I am not sure why, but I am not complaining. What is getting me down a bit, it the slow speed during the day (and then inconsistent increase during the late evening).
Let us know what happens tomorrow...
Have you tried to "flush" the dns (refer to my previous post whereby you go to the command prompt and type ipconfig /flushdns)? I have noticed a significant increase in speed and I have also had less drops in connection. I am not sure why, but I am not complaining. What is getting me down a bit, it the slow speed during the day (and then inconsistent increase during the late evening).
Let us know what happens tomorrow...
No thanks to Neotel, I think I may have found the problem. I spent the first day of my holiday waiting for, logging, and debugging the ppp disconnections from Neotel. It turns out they have a buggy ppp implementation. Here's a list of problems and workarounds.
1. It turns out they do not support the CCP (Compression Control Protocol) standard defined by the ppp specification. I added "noccp" to my ppp peers file and things improved a bit.
2. They do not handle the LCP (Link Control Protocol) standard defined by the ppp specification correctly. I set "lcp-echo-failure" to 0. At least this stopped my side from disconnecting due to LCP echo failure from their side.
3. I noticed the last procedure before the LCP termination request was a VJ-compression packet request from my side. I added "novj" and this resulted in less rejects and naks in to debug log. The link was still dropping regularly.
4. Finally after reading the pppd man-page I threw in "nomagic" and "noendpoint" for good measure. These parameters are recommended for buggy peers.
I have now been connected for 11 hours straight without a disconnect. Once or twice in the last three months I managed to stay connected for up to six hours. This is a new record.
1. It turns out they do not support the CCP (Compression Control Protocol) standard defined by the ppp specification. I added "noccp" to my ppp peers file and things improved a bit.
2. They do not handle the LCP (Link Control Protocol) standard defined by the ppp specification correctly. I set "lcp-echo-failure" to 0. At least this stopped my side from disconnecting due to LCP echo failure from their side.
3. I noticed the last procedure before the LCP termination request was a VJ-compression packet request from my side. I added "novj" and this resulted in less rejects and naks in to debug log. The link was still dropping regularly.
4. Finally after reading the pppd man-page I threw in "nomagic" and "noendpoint" for good measure. These parameters are recommended for buggy peers.
I have now been connected for 11 hours straight without a disconnect. Once or twice in the last three months I managed to stay connected for up to six hours. This is a new record.
Scratch that. At 1 second past 11PM I started receiving the termination requests again. I've had 82 disconnections in that past 9 hours.
Out of nowhere I receive a disconnection request. The link shuts down in an orderly manner. Then, before the link finally terminates I get an authentication request of all things. I've included a typical bit of log below.
02:13:42 : rcvd [LCP ConfReq id=0x2 <mru 1500> <asyncmap 0x0> <auth chap MD5> <pcomp> <accomp>]
02:13:42 : Connect time 3.2 minutes.
02:13:42 : Sent 305 bytes, received 1729 bytes.
02:13:42 : Script /etc/ppp/ip-down started (pid 30572)
02:13:42 : sent [LCP ConfReq id=0x90 <asyncmap 0x0> <pcomp> <accomp>]
02:13:42 : sent [LCP ConfAck id=0x2 <mru 1500> <asyncmap 0x0> <auth chap MD5> <pcomp> <accomp>]
02:13:42 : Script /etc/ppp/ip-down finished (pid 30572), status = 0x0
02:13:42 : rcvd [LCP ConfAck id=0x90 <asyncmap 0x0> <pcomp> <accomp>]
02:13:42 : rcvd [CHAP Challenge id=0x1 <xxKEYxx>, name = "pdsn"]
02:13:42 : sent [CHAP Response id=0x1 <xxKEYxx>, name = "[email protected]"]
02:13:42 : rcvd [CHAP Success id=0x1 "Welcome to pdsn."]
02:13:42 : CHAP authentication succeeded: Welcome to pdsn.
02:13:42 : CHAP authentication succeeded
02:13:42 : sent [IPCP ConfReq id=0x73 <addr 0.0.0.0> <ms-dns1 0.0.0.0> <ms-dns3 0.0.0.0>]
02:13:45 : sent [IPCP ConfReq id=0x73 <addr 0.0.0.0> <ms-dns1 0.0.0.0> <ms-dns3 0.0.0.0>]
02:13:47 : rcvd [LCP TermReq id=0x3]
02:13:47 : LCP terminated by peer
02:13:47 : sent [LCP TermAck id=0x3]
02:13:50 : Connection terminated.
02:13:51 : Hangup (SIGHUP)
02:13:51 : Modem hangup
Out of nowhere I receive a disconnection request. The link shuts down in an orderly manner. Then, before the link finally terminates I get an authentication request of all things. I've included a typical bit of log below.
02:13:42 : rcvd [LCP ConfReq id=0x2 <mru 1500> <asyncmap 0x0> <auth chap MD5> <pcomp> <accomp>]
02:13:42 : Connect time 3.2 minutes.
02:13:42 : Sent 305 bytes, received 1729 bytes.
02:13:42 : Script /etc/ppp/ip-down started (pid 30572)
02:13:42 : sent [LCP ConfReq id=0x90 <asyncmap 0x0> <pcomp> <accomp>]
02:13:42 : sent [LCP ConfAck id=0x2 <mru 1500> <asyncmap 0x0> <auth chap MD5> <pcomp> <accomp>]
02:13:42 : Script /etc/ppp/ip-down finished (pid 30572), status = 0x0
02:13:42 : rcvd [LCP ConfAck id=0x90 <asyncmap 0x0> <pcomp> <accomp>]
02:13:42 : rcvd [CHAP Challenge id=0x1 <xxKEYxx>, name = "pdsn"]
02:13:42 : sent [CHAP Response id=0x1 <xxKEYxx>, name = "[email protected]"]
02:13:42 : rcvd [CHAP Success id=0x1 "Welcome to pdsn."]
02:13:42 : CHAP authentication succeeded: Welcome to pdsn.
02:13:42 : CHAP authentication succeeded
02:13:42 : sent [IPCP ConfReq id=0x73 <addr 0.0.0.0> <ms-dns1 0.0.0.0> <ms-dns3 0.0.0.0>]
02:13:45 : sent [IPCP ConfReq id=0x73 <addr 0.0.0.0> <ms-dns1 0.0.0.0> <ms-dns3 0.0.0.0>]
02:13:47 : rcvd [LCP TermReq id=0x3]
02:13:47 : LCP terminated by peer
02:13:47 : sent [LCP TermAck id=0x3]
02:13:50 : Connection terminated.
02:13:51 : Hangup (SIGHUP)
02:13:51 : Modem hangup