Crystal Web (NEW ISP) ADSL Feedback

Status
Not open for further replies.
Looks like we need to do an entire reconfiguration of the support system - it may take a few hours. Apologies about this to all - we will be back online 100% later this afternoon or early evening.
 
Can somebody remove the restriction that limits the line to my telephone number please. What do you need from me?
 
Just signed up. Hope I'm not going to be disappointed. I still have my afrihost business account until the end of the month - hopefully I won't be needing it
 
Paul/Shaun

Two things quick:

Lovin t3h interwebs, but during the day get odd packetloss as we did before between your jhb and cpt cores:
cptpl.png


Second thing: Do you guys know when you'll have the dns pointer service running? (so that I can get home's IP by looking up user12345.dns.crystalweb.co.za or something similar)
 
Still very tough to diagnose properly when the exchange is reporting as it is in those results. :(

We will release the DynDNS service to all customers in the coming weeks, but for mybroadband members we'll offer it for a limited few for now. Just ask support and they will process it for you. At no additional cost.
 
Crystal Web said:
Just because switching shows a different result on hop 2 does not indicate that the issue is with the ISP. It quite physically cannot be. That is not our network, so it still indicates something that needs to be reported to Telkom.

Try to understand from our perspective when trying to offer support - it's a troubleshooting game - there isn't a switch that we can flick that says "no more packet loss", so we have to eliminate possible problems. When a glaringly obvious issue of packet loss exists, it's necessary to understand where it is happening, so tracerts and pathpings show this to us. When the issue starts at the second hop, while we'd love to fix it, we simply cannot. There's no mechanism for us to go and make changes to Telkom's network in any way whatsoever.

Ping response times are also not the best check, as response times may be okay, but packets are dropping, which is why we ask for better and more accurate pathpings or mtr tests. I'm quite sure Laurie could have written a book on the multitude of problems, but instead was asked a specific question and as with any form of journalism, had his answer condensed.

You must realise that we get as frustrated with this nonsense of congestion as anyone else. Probably more. We effectively have to spend our days diagnosing another company's problems, and spend ages trying to explain to customers the intricate details of congestion when it's spotted, and we're expected to fix it. It frustrates the living daylights out of me personally, as we want nothing more than a working last mile for everyone. Support could actually focus on our network rather than becoming an extension of Telkom's instead.

However when congestion or line-related issues are picked up, we will immediately ask for that to be rectified, or a test at another time to be performed without the congestion or line troubles. When something as fundamental as line or exchange troubles is present, it's nigh on impossible for us to identify any subsequent problems properly unless it too is blatantly obvious to us, in which case we'll sort it out.

Also remember that routing and networking don't happen at the account level, but at the wider network level. So it's highly unlikely that the issue is specific to your account - either we have a general problem, or in all likelihood Telkom has a general problem or one specific to your area, exchange, or line. It should also be noted that hops after the second are not all on Crystal Web's network. You'll often see the packets peer off onto another ISP's network - this is called peering. So while it's easy to look at a tracert with a novice's eye and think "my ISP has problems" it's often far more complicated than that, and requires deeper interrogation on a far deeper level. A trace can help to identify a possible problem - it doesn't necessarily mean that one exists.

My final word on this, is that congestion is a tricky beast to negotiate. It is by the nature of it, unpredictable and variable in extent. One minute it's congested; the next it's not. We do everything we can to help, and we actively avoid blaming an exchange as far as humanly possible. But it's necessary for one to eliminate that as a possible source of problems if it exists, before an ISP can properly diagnose any other non-glaringly obvious problems in the routing. If you say, my streaming is buffering, and we test and exchange congestion is identified, it's very difficult to move past that point and then attempt to diagnose something wrong with the account or the network in general. And if one were able to do so, it would be a very technical process, and not one that first line support could manage, as there would need to be an analysis of connection qualities on the various input and output stats we have, which requires some rather specialised skills and an understanding of how the packets are managed by the particular systems in place.

So I extend my sympathies to those with issues at their exchange - I really do, but it's not something we can fix, and not something we can overlook as being the cause of the problems.

Tl;DR analogy:

As a basic analogy, imagine a stream of water flowing from a tap, into a drain. You observe from down by the drain that it's dripping and not flowing properly, and ask a plumber to fix it. The plumber says something is stuck inside your tap itself causing the water to flow poorly, but it cannot be removed because the manufacturer sealed the unit. If you insist that the problem has nothing to do with the stuff stuck inside the tap, and ask the plumber to diagnose and fix other possible problems, he's in all likelihood not going to be able to help you much at all, no matter what he tries, and no matter what he tests, because the fundamental problem with the water flow is happening elsewhere. But you tell him when you switch on the hot water tap it flows better - there's no logical explanation for it - the plumber will still tell you to have the manufacturer fix the fundamental problem first before he can really move on, otherwise you need to replace it entirely.

That's about as basic but accurate an analogy I can think of right now to help explain the problems we face when asked to "fix" a customer's connection where clear evidence of a blockage in the tap exists. I really do hope that this helps to explain from our perspective why we ask you to run these tedious tests, and why it's incredibly difficult for us to "fix" things where the fundamental problem appears off our network. :)
Click to expand...

I do feel that this needs another posting as we've had a few support queries about this today, in particular where customers have said that switching accounts appears to solve problems at the Telkom exchange.
 
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%
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X