2.1: What is the definition of the term "Block Sync"?
Block sync is a technical term used in conjunction with the LANCity cablemodems. It is the condition
in which a LANCity is allowed transmission access to the local cable infrastructure. Block sync can be
lost if either your local upstream or downstream signal strength is too powerful or too weak. It can
also be lost if either your upstream or downstream channels are plagued by noise and/or interference.
The intermittant loss of block sync will cause packet loss, jitter (uneven packet latencies) and poor
upstream/downstream throughput. It is basically your worst enemy, with respect to cablemodem internet
access. The loss of block sync is indicated by a flickering or complete absence of the bottom of the
two, bright lights on the side of the LANCity cablemodem. Both of those lights should always be
completely solid and should never flicker or go out. If that is not the case, you definitely have a
problem on your hands. If the block sync light flickers or goes totally dark on a regular basis, your
connection is in serious jeopardy and should be corrected ASAP by Rogers Cable.
The block sync term, although originally used in conjunction with the LANCity, is often associated
with the Terayon cablemodems and many others. All cablemodems have some sort of "online" light on them.
This is the equivalent of the block sync light on the LANCity and should be looked upon with the same
importance.
3.1: How fast is Rogers Hi-Speed?
The maximum download speed with the Rogers Hi-Speed service is 1.5 Mbps (megabits per second), or
the equivalent of 187.5 KB/sec (kilobytes per second). Conversely, the maximum upload throughput
is 400 Kbps (kilobits per second), or the equivalent of 50 KB/sec (kilobytes per second).
Hi-Speed Lite Customers can only achieve a maximum download speed of 128kbps (kilobits per second), or
the equivalent of 16KB/sec. Maximum upload speed is limited to 64kbps, or 8K/sec.
That's what the service can deliver under optimal line conditions, data segment load and distribution/
backbone circuit load. However, a more realistic top end download speed, due to protocol overhead, is
generally in the 150 KB/sec to 160 KB/sec for Hi-Speed customers, and around 13K/sec for Hi-Speed Lite customers.
Of course, this doesn't mean that you will see that high a speed - there are many factors which
influence the effective throughput of your connection - and many of them are outside the control
of Rogers. Just as with any other internet service, cable is susceptible to widely varying
network performance depending on how busy the network and servers you're trying to reach, who else
is using the system at the same time as you, and what they are trying to do. As such, you may see
considerably less throughput than you might expect - even on a perfectly functional network.
That said, no subscriber should be paying a premium price for a less-than-premium service.
Unofficially we've learned that Rogers' Support people will try to ensure that Hi-Speed
minimum throughput speed, on a reliable test, exceeds 100 KB/sec (or roughly 50% of the theoretical
maximum). While we believe this information is reliable and acceptable, Rogers support personnel
continue to deny any knowledge of this guideline, even as they implement it.
3.2: Is there anything I can do to improve my network performance?
Well, you could talk your neighbours into one of the "other" services available, but that would
certainly ruin your midnight Quake games (and we wouldn't want that, would we?).
The first basic principle to apply is that your internet connection will not be quick, if your computer
is busy doing other things. Shut down all those taskbar "quickies".... they're just stealing CPU cycles
from you. Close down applications which are not in use. Don't overburden your computer with other
tasks if it's not capable of handling all of them. (I will point out that while it may be enjoyable
to listen to a CD while surfing the internet, doing so on a P100, with 32MB of ram, and using a
software-based wavetable synthesis sound card is almost certainly a death knell to low priority
tasks such as screen rendering).
Ok - now that you've got all the unneccessary programs shut down - how about your hard disk? When was
the last time you defragmented it? Surfing, by nature, results in a lot of information being written
to the disk cache for later reference. If your disk is fragmented, then you will find that it takes
longer to write the files to disk for caching purposes. Clean up some free space and defragment
regularly (this will help in many operations, not just with surfing).
Finally, there are registry changes which some people suggest will increase your speed. The most
common of these are the MTU (the Maximum Transmission Unit, or most amount of information
your computer is allowed to include per packet) and the RWIN (the receiving TCP window size,
or the amount of information your computer will buffer before requesting more). If statically
configured, I recommend that the MTU (Windows registry value of MaxMTU) be set to 1500,
with an accompanying RWIN (Windows registry value of DefaultRcvWindow) of either 32120 or
64240- no larger. If dynamically configured, I recommend that both the MTU and RWIN values be
completely removed and that MTU discovery be enabled. The relevant registry values are
PMTUDiscovery and PMTUBlackHoleDetect. They should be both set to 1. The dynamic
solution is preferred because performance-degrading packet fragmentation won't occur with this
setup. I am able to meet all the maximum performance parameters of the Rogers Hi-Speed service (i.e.
1.5 Mbps downstream, 192 Kbps upstream, <10 millisecond latency, etc.), provided that the proper
conditions exist, with the latter setup.
Please see this thread
at DSL Reports for more information on how to implement
the dynamic setup. Their tweak page contains links
to discussions on the RWIN and MTU values.
It must be kept in mind that some versions of Windows contain a default RWIN setting of 8192,
which will barely accomodate a 56K dialup modem. Due to this shortcoming, it is imperative that
the registry be tweaked if everything but the downloading speeds of your service seem fine. If you
are the curious type, checking your registry for the DefaultRcvWindow value is a good idea.
If it's set to 8192, either remove it completely and implement a dynamic setup, or tweak it according
to the above recommendations. And remember, if nothing is wrong with your connection and the speeds
are great, looking for more through registry tweaks probably isn't a good idea. Don't fix it if it
ain't broke!
First off, the LANCity is a dark grey colour, metallic and looks like a slightly-rounded rectangle
with ten little fins on top. The TeraPro comes in two versions: one that is jet black and rectangular,
and another version that is dark blue, which some subscribers liken to a shark fin. Both mainly consist
of a plastic construction.
The LANCity was one of the first kinds of cablemodems that ever appeared on the market, way back in
1995. Developed by Massachusetts-based LANCity Corporation, the company was soon bought out by Bay
Networks. In turn, Bay Networks was acquired by Nortel Networks in 1998. Sadly, Nortel chose to
discontinue the production of these impressive pieces of hardware after it consolidated Bay's
operations with its own. The LANCity architecture is based on the 10Broad36 Ethernet standard
and uses a mixture of both reserved and contention time slots for its MAC (media access control)
mechanism. Its physical layer makes use of Q-PSK (quadrature phase shift keying) modulation on two
different 6 MHz channels- one for upstream transmission and one for downstream. The LANCity is rated
to deliver a maximum of 10 Megabits per second in both directions, provided that it is not
rate-limited and in the proper environment. Symmetric broadband access products are a true rarity
nowadays, so this is definitely notable. The LANCity architecture is bridged, rather than routed,
and makes no use of any virtual circuit-based technologies like Frame Relay or ATM. Its design is
simple, elegant and efficient, delivering the maximum possible performance under a carefully managed
and maintained cable plant.
The TeraPro is among the "next generation" of cablemodems. It first appeared in 1999 as Terayon's
newest, full-featured offering to cable operators and their subscribers. The TeraPro architecture is
fully ATM-based, allocating a virtual circuit to each subscriber on the cable network. Since ATM is a
cell-based protocol, subscriber traffic can be managed with precision and data segments can be loaded
more aggressively. The TeraPro contains an extra feature, however, not found in most other cablemodems:
an Advanced "PHY", or physical layer. This feature is made possible by Terayon's implementation
of CDMA (code division multiple access), which is normally used in the newer digital cellular phones.
CDMA is a technology that spreads a signal, or carrier wave, over a particular range of frequencies.
By being able to spread a signal over a whole range of frequencies, the TeraPros can leverage the
cleaner part of the transmission spectrum against those parts that contain noise and interference.
For this reason, the TeraPro is one of the most resilient cablemodems out there when it comes to
combatting less-than-ideal line conditions. The rest of its physical layer makes use of 16-QAM
(quadrature amplitude) modulation on a 6 MHz downstream channel and a 5 MHz upstream channel.
The use of Q-PSK is an option on the upstream, depending on how noisy line conditions are. The
TeraPro has the ability to transmit and receive at 14 Megabits per second, provided, again, that
it is not rate-limited and in the proper environment. The LANCity and TeraPro may be the only
symmetric cablemodems that exist in the world right now. Although I am confident that the last
statement is accurate, I'm not 100% sure of its validity. If anyone can find another such device,
please let me know.
For more information on the TeraPro, please check here.
Technical specs no longer exist on the LANCity since it was discontinued, but I do have some
copies of those specs. I'll see if they can be posted here. Additionally, I wrote an (rather
technical) article comparing these two devices which can be found
here at
Cablecaster Magazine. Please have a look at it if
you are curious about this sort of thing.
4.1: Are there security issues when using Rogers Hi-Speed?
There are two types of security issues to be concerned about: one
is the Windows file sharing problem, while the other is the presence
of trojan horse servers. These problems, however, can affect not only
Rogers Hi-Speed subscribers, or even cablemodem users in general, but all
users who have Windows operating systems installed on their machines.
The internet service in question has absolutely no bearing on these
vulnerabilities.
Addressing the first of these, the file sharing problem is a
gaping security hole in the various Microsoft Windows operating
systems. It is the most notorious vulnerability of the Windows OSes.
This weakness can allow an outside machine access to all of your
computer's files. This is possible because the NetBIOS file sharing
ports are left open by default on all Windows machines. FYI, these are
ports 137, 138 and 139 and you can view them by doing a "netstat -na"
in a DOS window. It is up to the user to disable them manually. To do
this, you must go into your network configuration and remove all the
listed file sharing components. When that is completed, you will no
longer see those ports listed with the netstat -na command. FYI,
subscribers with LANCity and TeraPro cablemodems need not worry about this
issue since Rogers blocks, by default, all incoming and outgoing
packets to the NetBIOS ports.
Addressing the second issue, trojan horse servers are hidden pieces
of software on some computers that can allow unfettered outside access
to those machines. The best known trojan servers are Netbus and Back
Orifice. They affect Windows machines exclusively. You can only be
infected by one of these hidden servers by unwittingly opening an
executable file, designated by the .exe or .com suffix, which contains
the trojan server initialization files. For this reason, you should
never, ever open any executable attachments or downloaded files before
checking them first with the latest virus scanner definitions. You can
immediately see if you're affected at any point in time by doing a
"netstat -na" in a DOS box. This command will allow you to see which
ports on your system are listening, or open, to outside connections.
If you see ports 12345 (Netbus) or 31337 (Back Orifice) listening, you
can be sure you are infected with one of these servers. Luckily they
are fairly easy to disinfect with the appropriate programs. Fixes can
be found at SecurityFocus
or Packet Storm Security
by doing a search on either site. Also, if you see any port in the 0-
1023 or 5001-65535 range that is listening and, as far as you know,
have no apps running knowingly in the background, you may be infected
by one of the countless clones of these trojans. Again, you can do a
search to find out what you are infected with and then find the fix.
A program called Inzider by Arne
Vidstrom can help you determine what ports are bound to which programs,
thus allowing you to identify any trojans lurking on your system.
4.2: What is a firewall and should I be using one?
A firewall is a piece of software - or an expensive piece of
dedicated hardware - that controls access to various TCP and UDP
ports. With a firewall, an administrator can state which ports they
want to be accessible and to whom. For instance, a corporate admin
may only allow certain departments to get at certain services on their
designated machines. This may be done for security purposes (that is,
after all, the primary function of a firewall), or for other reasons
like internal traffic management. There are numerous firewalls out
there for home users concerned about their systems' security. ConSeal
is one of the better known personal firewalls, but there are many
others one can experiment with. A good firewall will allow the user to
state which networks are allowed to access resources on their system,
along with what ports may be accessed, and by which transport and
upper layer protocols. The more specifics included in the
configuration, the better, especially if you are intent on running any
casual servers. I strongly recommend against installing one of the
"dumbed down" firewalls, since they generally make it much more
difficult to block by specific port numbers and IP subnets. I
recommend running a firewall, not out of fear of an intrusion, but
mainly for curiosity's sake. They can tell you who is querying your
system and on what ports. FYI, the various Unix OSes, particularly
Linux, FreeBSD and OpenBSD, carry, for the most part, far superior
firewalls products to Windows. If you are very concerned about system
security, you might think about switching over to one of those OSes.
Ports that should be blocked right from the start are the NetBIOS
ports (UDP and TCP 137-139), along with Back Orifice (UDP 31337) and
NetBus (TCP 12345). If you are concerned about being ping flooded, you
should also block incoming ICMP Type 8 (echo request) and outgoing
ICMP Type 0 (echo response) packets. You should log any incoming
packets to ports 0-1023 and 5001-65535. Repeated attempts by
particular IPs on any ports, especially 23, 1243, 12345 and 31337,
should be reported immediately to the administration of the attacking
IP's domain. You can find out specific domain admin contact info by
doing a "whois" on the whois.arin.net database server. FYI, repeated
attempts on TCP port 119 by authorized-scan.net.rogers.wave.ca
(24.112.31.170), authorized-scan.security.home.net (24.0.94.130) and
authorized-scan1.security.home.net (24.0.0.203) can be ignored, as
they are routine scans for compromised NNTP news and SMTP mail servers.
Finally, it must be noted that subscribers with LANCity cablemodems
may experience what would at first appear to be constant intrusion
attempts. This is because the LANCitys make use of a bridged Ethernet
topology, which contains quite a bit of broadcast and multicast
traffic. Many firewalls won't filter out this type of traffic by
default, so it will seem like these packets are attempted intrusions
on your system. They are not, so there's no need to panic. To verify,
just check your logs to make sure that the destination address is an
address other than your own. If it's broadcast or multicast traffic,
the destination address will always differ from your own. For this
reason, it is important to configure your firewall to ignore (i.e.
disable logging and allow) any broadcast or multicast packets on the
network. Only cablemodems like the LANCity, which make extensive use
of broadcasting and multicasting, need to have their corresponding
firewalls configured in this manner.
5.1: What does the Rogers proxy do and should I use it?
The proxy is a local HTTP (web) server internal to a
regional/main data center, which caches (stores) frequently requested
content. It was originally implemented in order to greatly reduce
unnecessary network traffic, particularly backbone traffic leading to
the @Home Network in the US. When you want to access a website via
your proxy, you send a request to your proxy, which then checks to see
if any of the related content is stored locally. If it isn't, the
proxy will access the remote server and send back the information you
originally requested. The new information is then cached on the proxy
for a predetermined period of time.
The proxy is a good idea to use, particularly if your performance
is good. If performance is waning, then turn off the proxy and see if
it improves. If it does, you may want to avoid the proxy altogether.
The main thing to keep in mind here is that if every subscriber
avoided the proxy, a substantial amount of traffic would be placed on
the backbone. So please use it if the performance is reasonable.
Rogers Provides information for setting up the proxy
here - though it is for V5 of Internet Explorer. For V6, the only change
to the process necessary is that the second step is not necessary - you can
go directly to connections.
For manual configuration of the proxy, follow the steps mentioned above, but
rather than using automatic configuration, put a checkmark next to "Use a
proxy server", and use the Address of "http://cache" and the port "8080". (remove the quotes when entering).
5.3: What is "dynamic" setup?
A dynamic setup is described in the previous answer. It makes use
of the dynamic host configuration protocol (DHCP), to automatically
initialize your computer's network settings via a remote configuration
server. To repeat: enable DHCP (disable any static configuration),
specify your computer's name (in the format crXXXXXX-Y for Rogers,
where the X's are a number supplied on your invoice and the Y is an a,
b or c) and renew your IP via winipcfg (in Windows). To specifically
disable a static configuration, you need to go to TCP/IP properties
under your network configuration in Windows and disable WINS, DNS,
NetBIOS, all bindings, remove all listed gateways and set it to
retrieve an IP address automatically. FYI, when renewing your IP
manually, never, ever choose the "release" or "release all" options
under winipcfg. If you do this and your DHCP server fails, you'll
lose all network connectivity. Always choose the "renew" or
"renew all" options when manually renewing your IP.
5.4: What is "static" setup?
A static setup relies on your network configuration information
to be manually entered into your settings, as opposed to being
automatically initialized by a remote configuration server. To do
this, you need to enter the information originally supplied to you in
your invoice. You need to specify your IP address and its subnet mask,
your primary and secondary DNS servers, your default domain (in the
format segment1.on.wave.home.com, where segment1 is the name of your
segment), your domain suffix search order (again, in the format
segment1.on.wave.home.com), and add your corresponding gateway.
5.5: What are the advantages to using a dynamic or static setup?
This is a question that we hear a lot. In general, there really are
no advantages to using one or the other. If the DHCP server in your
regional data center is functioning properly, the use of a dynamic
configuration is strongly recommended. However, if it is not, a static
configuration is strongly recommended, particularly under Windows 98.
This is because Win 98 will automatically assign your computer an IP
address in the 169.254/16 class B address space, with a dynamic
configuration, in the absence of a DHCP server. In other words, if
you're using Windows 98 and your DHCP server fails, you won't be able
to get a proper IP and thus, won't have any connectivity. Win95 doesn't
have this problem because it will keep the last IP that was assigned
via DHCP. As long as you don't choose the "release" or "release all"
options under winipcfg, you'll be able to keep your IP and preserve
your network connectivity. Again, this is only a problem if your RDC's
DHCP server is down.
It is worth noting that, under the Rogers Hi-Speed service, not only
are DHCP leases 7 days in duration, but IP addresses are bound to
subscriber accounts. For this reason, your IP, although technically
dynamically allocated, may never change. As a matter of fact, my own
IP address is still the same on this very day as it was roughly 2
years ago, when I first had my service installed. FYI, the only thing
that must be kept in mind from time to time when using a static
configuration, is that IP addresses are renumerated on the rare
occasion. Although it isn't very likely that anyone with a LANCity
will ever have their IP change on them, TeraPro subscribers are more
vulnerable to such a transition. This is due to the differences in how
IPs are allocated under each of the two architectures. For this reason,
anyone with a static configuration must be fully aware that they will
lose complete network connectivity upon an IP address renumeration. If
this ever happens, you'll have to go back to a dynamic configuration
to get your new IP and then reconfigure statically using the new
address and subnet mask.
Think you've got a FAQ? Send it on over and we'll add it to the list!
n.b. creating an FAQ is VERY time consuming. For this
reason, we ask that you please bear with us while we continue to
improve, update and expand upon the information which is currently
present here. Our FAQ is a work in progress and is likely to remain
that way for an indefinite period of time. Fresh, new and thought
provoking questions are continually posed by many of you and we'll
eventually answer all of them in this section of our site.