Crystal Web (NEW ISP) ADSL Feedback

[Crystal Web]

Download :: 1.6 Mbps 204 kB/s
Upload :: 1 Mbps 128 kB/s

This on a 40Mb/s line...really????

This issue of mine has been with CW since the start of this month. Please can I get some feedback!

PS. This is for international only. Local is 100%

Spoken with the network team now. Can you hop on live chat with me quick and we'll arrange to sort this out for you.

 

[VScheepers]

Im on your live support chat

Spoken with the network team now. Can you hop on live chat with me quick and we'll arrange to sort this out for you.

 

[balky]

What exactly does this service do?

It effectively gives you a static IP address by using a URL. So you will be given a URL that always points to your IP address, even if your IP address changes.

 

[Ripstorm]

Hey Crystal Web,
Just got my CW account. Thanks for that.
One question though...uhhm..how do I pay? I have yet to give you any of my banking details

 

[Gatsby]

Hi Crystalweb,

I have a bit of a bone to pick with your support channels, I have kept this complaint off the forum for the past couple of weeks but I am beginning to run short on patience now. I have been visiting your live chat and emailing through numerous pathpings, tracerts etc. every time it happens. Tonight I was asked to run the tests again (6th time by my count) and I'm sorry but I just don't have the time to waste, an hour to hour and a half, every time I experience the same problems.

I have been told:

Congestion - It never is.
Routing Issues - Power cycles never help for long.
DNS proxy - same results with default dns and proxy dns. these have been sent through to support as well.
I have been told it has been escalated - never get any feedback, if I inquire as to the progress I get told they will inquire and get back to me, they never do.

I have run all tests back to back and followed them up with results from an afrihost account and all the issues described as nolonger present. I have sent through the the results from both CW and Afrihost run a few minutes apart and your support has conceded that the problem must lie on your side but it never gets fixed.

Please i beg of you, could I please get an update on what is happening. I see I am not the only one with these sort of issues....please investigate.

Looking forward to your reply.


EDIT: all of this started shortly after the switch from one IP range to the other.

 

[Ripstorm]

So, I've just tried the CW uncapped account. Can't really say that I'm impressed:

http://www.speedtest.net/my-result/3815143950

C:\Users\Ripstorm>tracert google.co.za

Tracing route to google.co.za [74.125.233.23]
over a maximum of 30 hops:

1 2 ms 2 ms 1 ms TP-LINK.Home [192.168.1.1]
2 12 ms 12 ms 11 ms 105.233.20.1
3 19 ms * 18 ms 196.6.121.38
4 16 ms 18 ms 17 ms 96.165.ct-dc.ctcore.net [196.41.96.165]
5 40 ms 39 ms 37 ms 196.6.121.237
6 39 ms 36 ms 39 ms pk101-jhb.jhbcore.net [196.41.96.157]
7 40 ms 40 ms 39 ms 72.14.220.38
8 48 ms 45 ms 43 ms 72.14.237.239
9 38 ms 39 ms 38 ms jnb01s07-in-f23.1e100.net [74.125.233.23]

Popped in my afrihost business account (which is usually horrendous):

http://www.speedtest.net/my-result/3815152267

C:\Users\Ripstorm>tracert google.co.za

Tracing route to google.co.za [74.125.230.248]
over a maximum of 30 hops:

1 1 ms 2 ms 5 ms TP-LINK.Home [192.168.1.1]
2 11 ms 10 ms 11 ms 105-236-5-1-esr-lo.mtnbusiness.co.za [105.236.5.
1]
3 12 ms 12 ms 12 ms ipc-send-tb-3a.mtnbusiness.net [41.181.53.214]
4 11 ms 12 ms 11 ms tb-dca-2.za--qux-a.za.mtnbusiness.net [41.181.18
4.28]
5 12 ms 12 ms 13 ms compj-cpt-1.mtnns.net [196.44.18.2]
6 157 ms 198 ms 158 ms ls-cr-2.uk--tb-cr-1.za.mtnns.net [196.44.31.113]

7 162 ms 156 ms 161 ms ls-pr-2.uk--ls-cr-2.uk-a.mtn.net [209.212.111.18
7]
8 157 ms 157 ms 156 ms rb-ad-1--msbm-rbg-1-a.mtnns.net [209.212.99.13]

9 169 ms 155 ms * 209.85.255.78
10 157 ms 156 ms 158 ms 209.85.250.169
11 156 ms 155 ms 156 ms lhr08s06-in-f24.1e100.net [74.125.230.248]

Trace complete.


Looks like this account isn't all it's cracked up to be

 

[Crystal Web]

Please try actual real world performance rather than speedtests and let us know what specifically isn't working for you.

 

[Crystal Web]

Hi Crystalweb,

I have a bit of a bone to pick with your support channels, I have kept this complaint off the forum for the past couple of weeks but I am beginning to run short on patience now. I have been visiting your live chat and emailing through numerous pathpings, tracerts etc. every time it happens. Tonight I was asked to run the tests again (6th time by my count) and I'm sorry but I just don't have the time to waste, an hour to hour and a half, every time I experience the same problems.

I have been told:

Congestion - It never is.
Routing Issues - Power cycles never help for long.
DNS proxy - same results with default dns and proxy dns. these have been sent through to support as well.
I have been told it has been escalated - never get any feedback, if I inquire as to the progress I get told they will inquire and get back to me, they never do.

I have run all tests back to back and followed them up with results from an afrihost account and all the issues described as nolonger present. I have sent through the the results from both CW and Afrihost run a few minutes apart and your support has conceded that the problem must lie on your side but it never gets fixed.

Please i beg of you, could I please get an update on what is happening. I see I am not the only one with these sort of issues....please investigate.

Looking forward to your reply.


EDIT: all of this started shortly after the switch from one IP range to the other.

Paul took over your chat earlier and I monitored it. Will have to PM you when I'm back online on the support system later this evening then as you told Paul that you didn't want to run any tests and then logged off, and neither of us are around to directly support now.

 

[Ripstorm]

Please try actual real world performance rather than speedtests and let us know what specifically isn't working for you.

Thanks for the response. I've tried twitch and it's buffering every 10secs on high. I do have the same issue with Afrihost (but hey, it's afrihost so that is expected). My free Web Africa test account is working flawlessly though

 

[Crystal Web]

Thanks for the response. I've tried twitch and it's buffering every 10secs on high. I do have the same issue with Afrihost (but hey, it's afrihost so that is expected). My free Web Africa test account is working flawlessly though

Personally I despise the way Twitch works and can't wait for them to move over on to Amazon's CDN nodes asap. But I did load up a Twitch stream and it was working.

To be honest, it's going to be tough for me to personally do anything right now, but live support are available to talk to if you'd like to see if there are any problems we can pick up.

 

[Ripstorm]

Personally I despise the way Twitch works and can't wait for them to move over on to Amazon's CDN nodes asap. But I did load up a Twitch stream and it was working.

To be honest, it's going to be tough for me to personally do anything right now, but live support are available to talk to if you'd like to see if there are any problems we can pick up.

I have to agree with you in that aspect. It's just strange that everybody on this forum are raving about how amazingly twitch streams on CW. For some reason I just don't see it. I'll have a chat with the live support and see what they can do.

 

[Intox]

I have to agree with you in that aspect. It's just strange that everybody on this forum are raving about how amazingly twitch streams on CW. For some reason I just don't see it. I'll have a chat with the live support and see what they can do.

Tried Twitch earlier today and it was buffering for me too.

 

[Gatsby]

Paul took over your chat earlier and I monitored it. Will have to PM you when I'm back online on the support system later this evening then as you told Paul that you didn't want to run any tests and then logged off, and neither of us are around to directly support now.

you left out the word "again". I don't see how running the same tests 6 times help when the results are the same every single time.

 

[Crystal Web]

you left out the word "again". I don't see how running the same tests 6 times help when the results are the same every single time.

I appreciate that to you they may seem like completely irrelevant and pointless tests, and that they appear to show the same things, but internally they don't, and are incredibly meaningful and helpful.

 

[FHTheron]

I like Crystalweb. I like the idea of Fusion Uncapped and and want Crystalweb to work. But I can't deny my perception of quality and reliability, so I went elsewhere. Hope to return here really soon when things are as great as they sound.

The images below were me doing this on a non-standard port:

ssh somehost 'dd if=/dev/urandom bs=1M' > /dev/null

sitting in Cape Town pulling from Centurion. My line only goes to 3.2Mbps, so I really want to get there all the time.




I asked support, who asked for pathpings. Over the next two or three days I was sending traceroutes from my Mikrotik after rebooting it and doing all the usual silly stuff. The Mikrotik traceroute is like mtr or pathping, so useful and shows packet loss. I was switching between PPPoE accounts and sending traceroutes to support to show that I get notable packet loss and VPN is useless. At that point Crystalweb on this forum said that account switching can't always be a reliable comparison for various reasons that I'll accept. So I started to keep three of my PPPoE accounts connected all the time, only letting one of them set DNS and default route. So, telling traceroute which interface to use, I was running three traceroutes at the same time using three different ISPs. Crystalweb usually showed packet loss to local servers (not always, but often enough to be a problem) like google.com, mweb.co.za and www.crystalweb.co.za. Another red ISP only rarely had packet loss, but never full speed and only one ISP never had packet loss with full speed nearly all the time - there are always issues everywhere, but you try to go where you see them least.

So that'll be my rant. I didn't really want to say anything bad, but this is mybroadband and this is an ISP that promotes transparency, proficiency and honesty, so I figured with everyone saying they have issues, I might as well put my personal perception on here too.

 

[Crystal Web]

Appreciate the feedback. May I ask why SSH on a non-standard port was the baseline test for your consumer account?

Also, on the VPN side of things, we prioritise them on condition that we recognise them - if not, and it's a legit VPN, we'll simply add the service to the list of recognised VPNs on the network and all should be good.

 

[Impregim]

From our CW Account. All looks fine - note!!! twitch.tv sometimes goes [Tits-UP]! So, don't blame your ISP.

Some more Relevant info:

PING bbc.co.uk (212.58.244.18) 56(84) bytes of data.
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=1 ttl=52 time=179 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=2 ttl=52 time=178 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=3 ttl=52 time=180 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=4 ttl=52 time=180 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=5 ttl=52 time=180 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=6 ttl=52 time=177 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=7 ttl=52 time=179 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=8 ttl=52 time=178 ms
64 bytes from fmt-vip72.telhc.bbc.co.uk (212.58.244.18): icmp_seq=9 ttl=52 time=179 ms
^C
--- bbc.co.uk ping statistics ---
9 packets transmitted, 9 received, 0% packet loss, time 8010ms
rtt min/avg/max/mdev = 177.448/179.297/180.468/1.084 ms

Host twitch.tv                         Jint   Last   Avg  Best  Wrst StDev
1. 105.233.15.1                        198.   41.2  13.9   8.6  93.8  15.8
2. 196.6.121.198                       151.   59.2  17.2  11.1  81.9  13.6
3. pk202-jhb-gi-2701.borwood.net       211.   73.3  16.7  12.1  88.7  11.8
4. th202-lon-po-sat3.borwood.net       743.  248.7 282.7 171.9 11219  1082.
5. th102-lon-vl-prcera2.borwood.net    519.  285.3 184.7 176.3 308.1  22.9
6. th302-lon-gi-1204.borwood.net       262.  198.3 180.9 176.0 301.2  13.3
7. ???
NetRange:       192.16.64.0 - 192.16.71.255
CIDR:           192.16.64.0/21
OriginAS:       AS46489
NetName:        JUSTINTV-2
NetHandle:      NET-192-16-64-0-1
Parent:         NET-192-0-0-0-0
NetType:        Direct Assignment
Comment:        http://www.twitch.tv
Comment:        NOC hours are 24/7
RegDate:        2012-11-05
Updated:        2012-11-05
Ref:            http://whois.arin.net/rest/net/NET-192-16-64-0-1

traceroute to twitch.tv (192.16.71.181), 30 hops max, 60 byte packets
 1  gateway (105.233.15.1)  9.638 ms  12.071 ms  14.012 ms
 2  196.6.121.198 (196.6.121.198)  19.674 ms  22.603 ms  25.544 ms
 3  pk202-jhb-gi-2701.borwood.net (217.78.94.146)  28.481 ms  29.942 ms  32.866 ms
 4  th202-lon-po-sat3.borwood.net (217.78.80.185)  196.449 ms  198.417 ms  200.863 ms
 5  th102-lon-vl-prcera2.borwood.net (217.78.88.26)  207.983 ms  211.425 ms  211.897 ms
 6  th302-lon-gi-1204.borwood.net (217.78.88.82)  213.837 ms  207.226 ms  206.787 ms
 7  * * *

traceroute to thinkbroadband.com (80.249.99.130), 30 hops max, 60 byte packets
 1  gateway (105.233.15.1)  10.232 ms  12.188 ms  14.627 ms
 2  196.6.121.198 (196.6.121.198)  20.775 ms  23.206 ms  26.143 ms
 3  pk202-jhb-gi-2701.borwood.net (217.78.94.146)  28.082 ms  31.028 ms  32.505 ms
 4  th202-lon-po-sat3.borwood.net (217.78.80.185)  196.075 ms  199.018 ms  200.485 ms
 5  th102-lon-vl-prcera2.borwood.net (217.78.88.26)  207.583 ms  209.063 ms  211.998 ms
 6  gi-0-3-0-1-63.r02.londen03.uk.bb.gin.ntt.net (213.130.48.1)  214.437 ms  208.237 ms  209.257 ms
 7  xe-1-0-1.mpr1.lhr15.uk.above.net (64.125.13.9)  208.785 ms  205.102 ms  204.822 ms
 8  94.31.40.154.IPYX-080002-001-ZYO.above.net (94.31.40.154)  204.808 ms  206.182 ms  204.814 ms
 9  te1-1-1-31.core-rs2.thdo.ncuk.net (80.249.97.85)  204.811 ms  176.106 ms  189.784 ms
10  * * *

Oh '' Good evening CW Guys

 

[FHTheron]

Appreciate the feedback. May I ask why SSH on a non-standard port was the baseline test for your consumer account?

Also, on the VPN side of things, we prioritise them on condition that we recognise them - if not, and it's a legit VPN, we'll simply add the service to the list of recognised VPNs on the network and all should be good.

That'll be the clincher yes, using a consumer account for business use. But I figured non-standard ports are a good test for unshaped data (the port is not under my control, company chose it). I had great youtube during business hours and normal hours, but Deezer always buffered when I wanted to use it (local MWEB server) which is probably also due to packet loss at those times. Legit Cisco VPN was also not working (default port, not even sure what it is to be honest) when there was packet loss.

I know expecting line rate on all ports at all times essentially means I should be leasing the bandwidth capacity directly at the real (high) cost, but everything said about the Fusion accounts made it sound like that would somehow be possible for the first 100GiB. Most users abuse uncapped accounts with non-stop downloading, I "abuse" any ISP account by doing atypical, but low-data, things.

I still have the month left on my account and I'm certainly going to keep trying it for things every day, but if you are going to launch business style accounts in future I'll probably look at those (as long as they are not just the inverse of consumer, but consumer++).

 

[Crystal Web]

That'll be the clincher yes, using a consumer account for business use. But I figured non-standard ports are a good test for unshaped data (the port is not under my control, company chose it). I had great youtube during business hours and normal hours, but Deezer always buffered when I wanted to use it (local MWEB server) which is probably also due to packet loss at those times. Legit Cisco VPN was also not working (default port, not even sure what it is to be honest) when there was packet loss.

I know expecting line rate on all ports at all times essentially means I should be leasing the bandwidth capacity directly at the real (high) cost, but everything said about the Fusion accounts made it sound like that would somehow be possible for the first 100GiB. Most users abuse uncapped accounts with non-stop downloading, I "abuse" any ISP account by doing atypical, but low-data, things.

I still have the month left on my account and I'm certainly going to keep trying it for things every day, but if you are going to launch business style accounts in future I'll probably look at those (as long as they are not just the inverse of consumer, but consumer++).

Really appreciate the feedback. Well the bandwidth is certainly still contended, that's for sure, so normal networking QoS takes place. You can understand that because of this, some protocols will always have to take priority over others. If this wasn't in place we'd see packet loss across absolutely everything. Importantly though, it's really a case of trying to ensure the best possible management of this, and I think for the price you pay it's pretty much the best quality account you can obtain in this respect.

We will certainly look to improve on certain services - Mweb servers we've seen some real trouble with. Internally it has sparked some rather frustrated debates about net neutrality as well.

 

[Crystal Web]

Oh '' Good evening CW Guys

[video=youtube;zGNVU5ZjlgA]https://www.youtube.com/watch?v=zGNVU5ZjlgA[/video]