Any idea what might be going on?
THanks,
David Jameson
Maybe the wan or an inside ethernet connection isn't negotiating to
100Mb/Full duplex. 5 Mbps would be about the limit if any ethernet
device/port was running 10Mb/half duplex.
D
D
2) I don't care how many features are involved - there should NOT be a
decrease in speed from 25Mbs down to 5 or 6Mbs just because of a
firewall.
Both of the above imply a very poor implementation. I simply don't
believe that firewall programs generally should cause that much of a
slowdown.
D
D
--->It would not be "of poor quality" it just means that with all the
features enabled that the unit is not ALSO capable of doing 100mbps -
Test 1) C <--> M Download speed = 25Mbs
Test 2) C <--> F <--> M Download speed = 5 Mbs
No other devices were involved in this test - therefore LAN to LAN
speed is irrelevent
----> LAN to LAN speed - do a test between machines INSIDE the lan
Yep - MTU is set to 1500 everywhere.
---> LAN to WAN Speed - have you checked the MTU settings to see if
that's
an issue from your old settings?
Agreed - but I already did these tests before posting on this forum.
--->The MTU setting could greatly impact your service level, and so
could
your choice of testing sites. MTU Settings run anywhere from 1430 to
1500 depending on the type of service, DSL needing lower settings,
Cable allowing higher settings.
Yep - that's why I bought it - and I didn't expect this problem to
arise.
--->The FSV318 is a good device, I've not experienced the problem you
have
when we use them to isolate lan segments.
The WAN port of the FVS318 is 10mb base-T.
E.
Oops, the v3 is 10/100.
E.
Oops part 2: Up to 11.5Mbps WAN-to-Lan throughput, 2.1 MBps 3DES throughput.
E.
--->Oops part 2: Up to 11.5Mbps WAN-to-Lan throughput, 2.1 MBps 3DES
throughput.
>From a cursory Google search for "firewall throughput", I'm seeing that
typical WORSE case download speeds when using firewalls are at least
60Mbs.
So is the Netgear a dud? Or is there some hidden adjustment that needs
to be made to it?
Thanks,
D
--->HA! I found it! If I have Keyword blocking turned on my bandwidth
is limited to about 190k. All I did was turn off Keywork blocking and
remove the one web site I'd blocked (myspace.com) and I have the full
speed of the cable back.Yet another problem with Netgear.
The 11.5Mbps Wan-to-LAN throughput (i.e. maximum download speed) was
straight out of Netgear specs for that model. I guess Netgear's
documentation is bodgy too.;-)
Good to hear you got it sorted tho.
Cheers,
E.
I appreciate all the responses from people. I definitely will not be
recommending Netgear to anyone anymore.
I wouldn't slam them too hard. It's advertised as a broadband router.
For the vast majority, broadband means 1.5 or maybe 3 Mbps. Connections
like your's are a relatively new phenomenon on the market. I only wish I
had your problem. (He says from behind a 512K satellite link...)
--
Rod
The WAN port supports 100 Mbps because a) 10/100 NICS are cheap as dirt,
and b) 11.5 is bigger than 10.
--
Rod
--->I wouldn't slam them too hard. It's advertised as a broadband
--->You really need to get a Usenet client that properly quotes people,
you're sucks (forgive the wording).
See my response from "Sat, Mar 25 2006 10:46 pm" where I indicated
quite clearly (I thought) that the MTU was set to 1500 everywhere.
--->If you didn't adjust the MTU setting, while is in the instructions,
then
you might not be getting what you need - you also didn't say if you
adjusted it or not in your reply.
Don't be condescending. I already pointed out that although the feature
was enabled, there were no actual keywords to test against. Even if
there were, this should not cause 20% degradation in throughput - think
about how the algorithm should be working - if it's set to block a
particular website, then it's going to check the OUTGOING http request
for a match --- that shouldn't have ANY significant impact on the
speed of the INCOMING data.
--->As for bad software, consider this, any time you enable a filtering
feature it has to execute SOME code, that code takes CPU Time (you do
understand that the device has some form of CPU, right?), and that
means
it will decrease performance for other things.
--->Why the heck would you assume anything - the specs are right on the
vendors site, please learn to read them before you make another
mistake.
Understood - but even the cheap sonicwall boxes are claiming at least
60Mbs throughput and they're not that much more expensive than the
Netgear devices.
--->Firewalls don't process at wire speed, they have a defined rate
that
they can process traffic at, it's been that way for ages.
One doesn't always have time to sit down and research the entire
firewall market. My assumptions seemed reasonable - I suspect a lot of
people will fall into a similar trap - in the worst case, this
newsgroup thread will be helpful to others.
--->Why did you not understand that you need to read the through-put
specs on any firewall
to determine if it's going to meet your needs.
Obviously - now that I know!
--->Before you make another choice, read the detailed specs to see what
the
throughput rating is, and it's not the port speed, it's listed in the
specs.
I appreciate your feedback - I'm off to look at higher end firewalls.
D
Sure I do - at least to some extent- but I simply didn't expect
performance to be impacted THAT much. I'm not (nor to I want to be) a
firewall expert. "Cheap" is a relative term - I remember when a 300
baud modem cost thousands of dollars - now a broadband modem costs $50.
Cheap does not imply subpar.
--->Then you don't understand features and how the impact performance.
Like
it or not, it's a CHEAP device and you should expect some trade-offs vs
a real firewall.
Thank you - it must be nice to be you - all-knowing and so forth. I
guess I'm just not as smart as you - I hope there's still room for me
on the planet!
I wonder where in the store I could have read the throughput spec? The
very detailed product specs on the side of the box (which I did read in
the store, taking more than 5 minutes) didn't include the throughput
number - otherwise I would have noticed it immediately. Is it your
conjecture that people should never go to a store for anything without
having done a detailed analysis in advance? Where is the time for such
analysis? The good news is that the store will take the product back.
--->Your assumptions are not reasonable, they were ignorant and ill
informed. If you took 5 minutes to read the specs on the device you
were
purchasing you would have seen the real performance stats.
D
BTW: I just found out the router (Cisco 1601r) that is before this thing
has a special serial connection that is limited to 2mbps, so when I
switch from the half T I use now to 4mbps dsl, I have to change this
router also!
Now I just found out the 1605 I was going to use also has a limit of
2mb/s!! I guess I need to get a better one!!
Here are some links, the router performance one is esp good!
Cisco_806
(http://www.cisco.com/warp/public/cc/pd/rt/800/prodlit/806bg_ds.htm)
http://www.cisco.com/warp/public/765/tools/quickreference/routerperformance.pdf
gr