Juniper vs Riverbed

W

Wong

Well-Known Member
Joined
Feb 4, 2009
Messages
278
Reaction score
0
Anyone here have any xp regarding Wan optimizing using these devices ?
 
Riverbed is very very good.
What kind of stuff are you looking at doing?
Pm dolby regarding riverbed information
 
Getting 2 Riverbed test units this week! Hoping there is a noticeable improvement on our DBN-CPT diginet line. I'll post back here when they're set up.

I really hope they're the proverbial ****. Sales guys and account managers are pushing these like theres no tomorrow. Honestly, I'm on the fence. Part of me thinks it will be an improvement, another part hardly an improvement, and another part feels like the improvement wont justify the cost. And no one has really touched on how well it performs with Exchange and Outlook, all they say is "It will be better". Better how, I'm not sure. I'm dealing with a lot of Online Exchange users and all mail is read in "real-time" from our Exchange server.

Should the testing go well I'll be looking to implement them at another 2 locations within the WAN.
 
Last edited:
Basically a Riverbed unit will almost double your throughput. IE: 1Meg diginet will become 2Meg. Obviously being point to point.
 
Grep said:
Basically a Riverbed unit will almost double your throughput. IE: 1Meg diginet will become 2Meg. Obviously being point to point.
Click to expand...

Thats a very broad statement? It depends on what you are using those links for doesnt it?
 
We are thinking of something similar to hungrybeaver 1 in cpt 1 in jhb and 1 more in the NOC - (not sure if im allowed to ask) hungrybeaver who did you get your riverbed units from and what did you pay for them ? We were quoted around R90k for 3 units does that sound about right ?
 
I'll need to confirm but I think they're from Networks Unlimited. Yep R90k for three sounds right. We were quoted R60k for two.

I'm trialling them for 2 weeks before I make a decision about purchasing them.

Make sure to get a trial unit first!
 
tau1z said:
Who is setting up these devices?
Click to expand...

A colleague of mine who has done it before.. I just don't have the time atm. Planning on doing a lot of tinkering though!

Still waiting for the units. I've been told to expect them on Thursday/Friday.
 
hungrybeaver - how do you find performance of the Riverbed units?

PS : Thanks for the lead taulz ;)
 
I really cant see how people can decide on what technology they must look at if they do not have a clear understanding of what is wrong on the network?

How do you really know that you need optimisation? - Is the line slow? Users complaining about performance? What do you base it on?

For example:

A client, where our products is installed, wanted to upgrade the line because they were complaining about slow application response time.
We quickly showed that the app was not running in the correct QOS and also showed that upgrading the line would not help.
The client phone his service provider that told him but it is in the correct QOS but in turn he could turn around and show his service provider that it actually was not.
Just putting it into the correct class we were able to reduce the response time from 20mins to to a few seconds.

Now he would have wasted money on upgrading the line and or getting optimisation.
(We are actually showing them the ROI of the Riverbed kit that they purchased, so another point we are helping them on)

So How does it work?

Client wants optimisation installed on a few sites that he knows is slow, but still he does not know what is the cause of the break in response time.
Where do you start? Because end user response time can be broken up into Application response time (Server think time), Data Transfer time (Time it takes for the data to go over the network) and then also Network Round Trip Time (Latency on the network) you can quickly make a decision on where to start. If you see high response time in Application response time you know that you must start looking at the server/apps, high Data Transfer Time, yes optimisation will work here, Network Round Trip Time, start doing investigation into the LAN and then also the links (Are they saturated perhaps?)

So if the client wants optimisation installed on a site that has high Application Response time it would be useless and he just lost money because his homework was not done.

There is all these things that people does not take into account.

Now I am not saying optimisation is a bad thing but how do you know it will really work? From only the reports that the specific Vendor will give you or a 3rd party monitoring solution? And it doesn't just stop there you will also be able to see if the proof of concept was a success because you will have a baseline before, during and then also after the POC.

Also which optimisation will be better for your company? Riverbed, Cisco WAAS, Bluecoat? There is a number of these guys that does well in their specific area. Like Bluecoat will give you better optimisation on HTTP than Riverbed but then Riverbed can do HTTPS better that the other guys.

//Thread jack over
 
From a non-technical side, a network admin may notice issues on his network. This could be his own logic, user complaints about speed, monitoring tools or other sources - but he'll know something is wrong and may have an idea of what it could be. Most system admins have heard of WAN optmization and may contact the 2 or 3 of the larger vendors to see what can be offered.

Each vendor contacted comes in and install 2 or 3 units at the sites chosen for a period of 2 weeks - at no charge. During this time the system admin can run a variety of his own tests and reports to get an idea of if something is happening. He can chat to users and get experience, send large files to other branches and pretty much anything he wants in that time. At the end of it, he also gets a pretty looking report with everything that happened during the POC - what was/wasn't optimized, dates/times, peaks, types, percentages etc

He compares all three of the big guys and - armed with pricing - can decide which works for him on his own network. If we're not selected - no harm done

In most our POCs, we're up against the competition and in 7/10 cases we're selected for the installation. A quick Google or check of the Gartner report will confirm our position in the market and performance ... and optimization of 40-60% isn't uncommon. We've done installs and due to deferring of a line upgrade, ROI was just under 12 months.

We obviously don't win every deal as there are other factors (ie price charged, budget available) and sometimes the traffic just isn't there. The optimization - even if it is there - may not be sufficient to warrant the extra expense of adding devices.

If anyone has the budget and interested to see what it can do for their traffic with a POC, just PM me and I'll see what I can do
 
Last edited:
@ snor - :D im only the jnr at the company I work for and I end up having to figure out stuff on my own most of the time , besides that we have done tests on our network trying to figure out how to improve performance with our web based applications we had vendors come in as well as telkom to perform their own network tests to confirm our finds - most of the vendors suggested we get either get juniper of riverbed to improve performance - as far as our lines they dont need to be upgraded nor do we need more bandwidth - a lot of our traffic can be thinned out using either of these 2 tools

We are in the process of getting a trail - always try before you buy :P
 
Dolby said:
From a non-technical side, a network admin may notice issues on his network. This could be his own logic, user complaints about speed, monitoring tools or other sources - but he'll know something is wrong and may have an idea of what it could be. Most system admins have heard of WAN optmization and may contact the 2 or 3 of the larger vendors to see what can be offered.
Click to expand...

I see what you are saying here and yes they might know where the problem is but most of the time he cannot pin point the problem. He/end users might be saying that the application is slow and then the techie assumes that it is the line or the server. In actual fact it could be the code that is defective and instead of doing a specific search on a database the developer might have screwed up a bit and put a value of search the whole database?

Each vendor contacted comes in and install 2 or 3 units at the sites chosen for a period of 2 weeks - at no charge. During this time the system admin can run a variety of his own tests and reports to get an idea of if something is happening. He can chat to users and get experience, send large files to other branches and pretty much anything he wants in that time. At the end of it, he also gets a pretty looking report with everything that happened during the POC - what was/wasn't optimized, dates/times, peaks, types, percentages etc
Click to expand...

Once again the customer chooses the sites that must be installed on and not you. How do you know that the client knows where the problem is? Wont it be a lot easier for you to actually tell the customer where to install the devices to get the best result out of the POC and ultimately show the customer the best ROI possible?

A key thing that you mention here is He can chat to users and get experience
We have seen multiple times over where the network is improved and then the users start complaining about the speed again but in actual fact the speed is still the same when they stopped complaining, they have just gotten used to the speed now.... :o This we will be able to show you how because we look at the actual traffic traversing the network and not synthetic data. We also look at the actual transactions and do DPI on most of the protocols to give you a deeper insight into the problems.

Also with our products we can help prove that your Optimisation is actually working because we do the baseline before, during and then after the POC and just show how mush value this tool can be for the customer.

We have done this for one of our customer's where they were running a POC on the big 3 and ultimately we showed which kit was the best for the organisation. Although everybody had their own pretty reports the customer could make a very easy decision based on what we gave him. FYI Riverbed won at this customer.

He compares all three of the big guys and - armed with pricing - can decide which works for him on his own network. If we're not selected - no harm done

In most our POCs, we're up against the competition and in 7/10 cases we're selected for the installation. A quick Google or check of the Gartner report will confirm our position in the market and performance ... and optimization of 40-60% isn't uncommon. We've done installs and due to deferring of a line upgrade, ROI was just under 12 months.
Click to expand...

We can discuss Gartner in dept but the forum might crash from the big post....
We obviously don't win every deal as there are other factors (ie price charged, budget available) and sometimes the traffic just isn't there. The optimization - even if it is there - may not be sufficient to warrant the extra expense of adding devices.
Click to expand...

That is why you need our tools to see if optimisation will work in the first place. :) Or where the money can be spent to have the best return for business.



This is not an attack on you Dolby or Riverbed. I am just bringing out the points that people does not think about. Like I said we work in conjunction with most of the optimisation guys so we can give the customer the best reports for his business decisions and if IT is improved, how it can also improve the business.

This tool also brings all the Silo's of business together, i.e. Networks (LAN,WAN), APP/Developers, CIO/CTO and business delivery.

So if anybody would like some more information on how we can improve your IT and ultimately make your business more money give me a shout.
 
We had a Riverbed test unit and work and all i can say it boy did that device help us. I used to get calls constantly about Email issues / Duplicates etc and since we had the Riverbed device i never got any calls @ all.

I used to monitor traffic on the Network and it improved tremondously. I was really impressed with that device.

Awesome Product !!!!
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X