Rogers Denies Blame For TekSavvy Outage Issues
As I noted last week
, indie Canadian ISP TekSavvy has been struggling with prolonged disconnections for many of their users, something the company says is because of changes at Canadian incumbent Rogers. To hear TekSavvy tell it, the company suddenly and inexplicably found their install and repair trouble tickets being ignored by Rogers; months of this contributing to a backlog of support issues that have caused massive headaches for the company and customers alike.
I asked Rogers for comment, and the company denies doing anything to intentionally sabotage TekSavvy's business. They also take issue with the number of impacted customers.
"Over the past several weeks we have seen an increase in the volume of issues and have increased our resources to resolve any outstanding issues as soon as possible," Rogers spokesperson Mary Pretotto confirms. "Although it has been suggested that there were 2,000 customers out of service, that figure is not correct and does not correspond to information we received directly from TekSavvy," she notes.
While TekSavvy CEO Marc Gaudrault stated
that unspecified Rogers changes were to blame for the hiccups, Rogers also denies that anything has changed.
"There have been no changes in processes with TekSavvy or in the network relationship," the company tells me. "All network improvements and enhancements have been part of normal operations."
In fact, the Rogers response to my inquiry lays most of the blame for the problems at TekSavvy's feet, claiming that a combination of "inaccurate" forecast data provided by TekSavvy and the everyday lag created by dealing with a third party provider are responsible.
"All third-party internet service providers submit a forecast to ensure Rogers allocates resources to meet projected demand," said Pretotto. "Unfortunately TekSavvy provided an inaccurate forecast that significantly underestimated their volume of traffic and overwhelmed the resources that Rogers has dedicated for third party internet providers."
As we saw in great detail with the CLEC/ILEC battles of old here in the States, Rogers has a vested interest in things not going particularly smoothly for a competitor. As such, it will be up to the Canadian regulators the CRTC to sort through the truth and determine if any foul play was involved.
Meanwhile, TekSavvy CEO Marc Gaudrault has posted a new blog entry
noting that things have improved slightly, but that the company is still working with lawyers and the CRTC to get things back to normal.
Both are to blame Listen, this is a combination of the old skool dragging their feet and techs/etc (union no less) doing their thing which is to mean up to government standards (which is D grade), AND TekSavvy being green by having customer service growing pains and a painful move to D3/ATPIA which IMHO they should have ramped up in more structured fashion. In fact my parents are still on old TPIA speeds, however they are now showing up under a slower ATPIA tier. Moving to D3 requires better signals and multiple channels, and guess what marginal signals are going to crap out when upgraded...
Add that and a dose of flooding and you have a sh**storm.
There is all reason to believe that Rogers process and TPIA in general sucks, as anything the government touches will have that glow, BUT Teksavvy has stumbled on their own two feet also....
Also, if you look at the other Rogers/TPIA forums, NONE are having the issues at the level as TekSavvy so logically concluding the blame is on both parties. Simple s**t like not having enough DHCP addresses at the CMTS is pure crap. The TS/Bell (DSL) seems to be better, so I think that both are fueling the fire and plain and simple is shows that Bell is a better operator than Rogers because they have been down the regulation road before... The D3 migration in general seems to be much worse than in the states...Not sure why but when TWC upgraded in my rents area, no issues and they are now running under Earthlink (quasi-TPIA).
North York, ON
Re: Both are to blame The problem at Rogers isn't the techs - they have been able to have a tech roll out to fix a Rogers customers' TV in under 4 hrs. The problem is the people at Rogers answering the tickets - or not answering them as the case seems to be.
Rogers doesn't inform TPIA prior to a node split, or when doing the upstream channel changes that they're doing currently (16QAM->64QAM) on a node, and when these changes are made, DHCP issues tend to occur. Because Rogers doesn't inform TPIA beforehand, and doesn't ask for additional IPs beforehand, TPIA experiences DHCP issues.
Hell, Rogers is even denying they're doing the change from 16QAM->64QAM, and bringing it from 4 upstream channels bonded, to only 3 - 2x64 & 1 16/64. Meanwhile, TPIA customers can see this quite clearly in their modem signals page. Its pretty ridiculous because Rogers' spokesmen are being fed lines of bull and spreading it to the media.
But hey, this is Rogers.. Providing bull to media is NORMAL. Remember the throttling issue - "Oh, no. We're not doing this." "Oh wait, yes we are, but it's not affecting anything else." "Oh, the CRTC's own tests says it is affecting other things. We'll stop throttling."
Battle.net Tech Support MVP
Re: Both are to blame I digress. The first time Rogers came out, they spliced my parents cable with the neighbors, and ignored the new line that was run two months prior when I went ballistic on the 1982 cable that was there and dropping whenever rain hit ground.
So a week later they come out, and hook up the original "new" cable and they missed the window.
So I'm not sure how that can be considered good practice or even right, but two calls and a lot of wasted time here.
I had a few PMs on DHCP and the issue (explained to me) was that Rogers assigns DHCP centrally, and TPIA through blocks at each CMTS. I'm not sure how fanning out to nodes would cause DHCP issues, but certainly changing modulation/bonding can cause signal issues. Apparently TS said they had no way of knowing they were running out of addresses, and it would take Rogers days to download the new profile to the CMTS and the feedback mechanism was people getting locked out of a new lease. That seems plausible, but if that is the case that is entirely dysfunctional...
So I'm assuming that ATPIA (I don't know) has say TekSavvy switches (10GB or the like) in the headend and that plugs into the CMTS or hub and then off to their network carrier to provide transit? I never saw how this is done, but would be interesting to know how they implemented this. So this means the CMTS still provides network services (DHCP/Docsis), just routes the assigned IP when seen at the CMTS to the Teksavvy "aggregated" switch pair (or more).
| |AkFubarAdmittedly, A Teksavvy Fan
. Typical Rogers response. Sorry but I believe Teksavvy on this. The ugly truth will come out when the CRTC deals transparently with this mess.
BHell... A Public Futility. When life throws you lemons, make lemonade.