Sunday, December 6, 2015

The Reality of 100 Mbps of Internet Service in Bangalore.

Internet is a very important service that anyone needs in today's world.

Thus, there is a constant effort that everyone makes to get to a better speed of Internet service. For example, it is well known that Internet-based videoconferencing, Cisco recommends, will need at least 384 kbps. And, for video streaming, Netflix recommends 3.0 Mbps for SD and 5.0 Mbps for HD quality.
So, when I saw an ad last year for a 40 Mbps Internet service, shown at right, I influenced a Bangalore-based relative to try that. Little did I realize what it'd mean in terms of actual Internet service experience.

When, eventually, the relative got a 100 Mbps from the same Internet Service Provider (ISP) recently, I had an opportunity to conduct some experiments on this new service. (These days, this ISP even markets a twin load service, a claim that says that both download and upload bandwidths are similar).

Here are some typical performances. Without even a ISP-supplied router, here's what I found as the various numbers at a Ethernet-attached Windows laptop directly to the ISP:

Source
DestinationDownload (Mbps)Upload (Mbps)
J P NagarBangalore87.5243.65
J P NagarSingapore13.108.47
J P NagarLos Angeles9.249.03
J P NagarSeattle7.4713.53

Thus, unless the ISP has good peering agreements with other global ISPs, the speed experienced at any given laptop for many international destinations will not be as claimed, even for globally important sites such as http://www.google.com. This is the experience that a user faces when connecting directly to the ISP, without a router in between.

Now, what happens when you introduce a home router between the laptop and the ISP? With the ISP-supplied home router, a different set of numbers is seen at the same Ethernet-attached Windows laptop:

Source
DestinationDownload (Mbps)Upload (Mbps)
J P NagarBangalore13.137.15
J P NagarSingapore1.390.45
J P NagarLos Angeles1.370.51
J P NagarSeattle0.990.50

Thus, with the introduction of the home router between the laptop and the ISP, the speed experienced at any given laptop for many international destinations further suffers.

Summary
1 Mbps download to certain destinations on an advertised 100 Mbps service? 0.5 Mbps upload? The actual realized speeds can be much less than the advertised speeds. In this example, they are 2 orders of magnitude inferior, with the ISP-supplied router, for specific destinations. What can we conclude from all of this?
  1. Unless the ISP has good peering agreements with global ISPs, the speed experienced at any given laptop for many international destinations will not be particularly good, even for globally important sites such as http://www.google.com. 
  2. The home router supplied by the ISP is also to be looked at. Just because the home router is supplied by the ISP, it doesn't mean that you can experience a good connection.
So, as they say, caveat emptor: Understand what you are getting into, before you sign up with any ISP.

2 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. I had occasion to make another experiment about a year later, with a Mac connected wirelessly to the ISP-supplied router in between the Mac and the Internet: http://www.speedtest.net/my-result/5903160127

    About 6 Mbps download, and 12 Mbps upload.

    In other words, there seems to be improvement in the service compared to a year ago.

    ReplyDelete