cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post
Level 7

WAN Optimization Reports Help

Hello:

We have Riverbeds deployed in our network and I downloaded the WAN Optimization reports module from this site to monitor their performance.  I'm wondering if anyone's had experience with this tool and can give me some pointers as to how to glean the best data.

So far, it seems as though monitoring traffic from my inpath interface to my next hop toward the internet (firewall) interface produces somewhat satisfactory data. However, I'm not sure I"m looking at the right thing. 

Can anyone share their own experience?  What interfaces did you monitor to get the best results?  How do you interpret the data?  What's the difference between pre-bytes/post-bytes and pre-packets/post-packets?   They seem to report very different results.

Thanks-

Leslie

0 Kudos
18 Replies
Level 8

i'm working on a similar issue. I am trying to report on the Percentage bandwidth reduction over time. I found that the Agg on the LAN resets around 4GB. this is because the values are stored as counter32 which means the value is wrapped when it hits 32 bits (4294967295).

The only way I see of being able to do this is if there was a way to transform the figure based on the previous poll, for example the new figure would be (newValue-previousValue)+previousValue. essentially incrementing by the differnence in the value and stoppin git resetting to zero.

Personally I think that Riverbed should set the device so that if either the LAN or WAN wraps and resets to zero then they both should.

I note that this post is quite old so if anyone has since had success in doing similar to what i'm trying to achieve i'd appreciate it.

0 Kudos
Level 8

Hello,

 

i am experiencing similar problems with negative values and was wondering if there has been any further developments? this affects both version 6.0.1 and 6.1.0 devices. i have run the cable swap tests and all appears to be OK. any suggestions please?

 

0 Kudos

Ross,

Do  the other Optimization percentages match what your management console is reporting? Does this appear to be a transient condition or does it stay at that incorrect negative percentage? We are just reading the values via SNMP so we are returning what the device is reporting. I will check to make sure those values haven't been deprecated.

Mav

0 Kudos


Having the negative number issues here as well. Checked the data from Orion against the actual riverbed data and we see the RB is accurate, Orion is hokey. Not sure what we have done wrong, just downloaded the reprot and ran it. Anyone want to help a struggling noob out here? Will continue reading but to be honest I am not very strong in Orion or the RB devices.

0 Kudos

Guys,

I have been working with tech support for quite a while on this now (case 448539)& so far we have come up with a couple of things.

#1 I had mistakenly been using my inpath0_0 for my "pre" interface. A call to Riverbed cleared that up, it was lan0_0.

#2 SolarWinds asked me to change the Pre & Post interfaces.

Now the "Pre" interface is the Rvb Wan0_0 & the "Post" interface is the Rvb Lan0_0 for all the remote locations.

See below thread for a discussion & attached images.

http://thwack.solarwinds.com/thread/25568


0 Kudos
Level 12

Ok, I now have positive optimization.  I just put the lan0_0 ID in the wan0_0 optimizedinterfaceid column for all my remote offices.  Then for my HQ that they all connect back to I did the opposite.  I put the wan0_0 id in the lan0_0 optimizedinterfaceid column.  I would think this would give me the optimized bytes and percentage from the riverbed unit.

0 Kudos
Level 12

I'm also just looking into this report and am confused from the guide instructions.  Mine don't have a firewall before the router, so I am looking at step 8. 

When it talks about the Inbound WAN interface for the router, is it wanting the actual router interface or the interface of the steelhead appliance?

Same goes for the Inbound LAN interface, is it wanting the interface on a switch or the lan0_0 on the stell head appliance?  I would think you would get the same data from either.

 

Right now I am getting negative optimization on the reports, so I am guessing I have it messed up.  Here my simple layout

 

HQ core switch -- riverbed -- router -- mpls cloud -- router -- riverbed -- remote office core switch

 

Are the inpath interfaces then not used for these reports at all?

 

Thanks

0 Kudos

It is wanting the lan facing and wan facing NIC ports on the Steelhead appliance for the report.

The inpath0_0 is the virtual representation of the physical connection between the lan0_0 and the wan0_0.  The report uses these two interfaces to calculate what comes in and out and then can use these numbers to calculate optimization.

I tried the router interfaces once, but it didn't seem to give me good data.

Below is an image of one of the reports I set up that gives me a good look at RB Optimization covering the last 24hours, last week, and last 30 days.

Very informative once you get it set up and have it running properly!

0 Kudos

Well I'm not sure what is different, but in order to make mine +%, my remote offices pre interfaces are the wan0_0 and my HQ is lan0_0.  So I don't understand how you guys are getting this to work with your pre interfaces being lan0_0.  Our units sit inline, not sure if that would make a difference.

0 Kudos


It looks as if you have your lan0_0 and your wan0_0 plugged into the wrong sides on most devices.

If you look at your I0-WA-2020-01 device, it is using lan0_0 in the pre and wan0_0 as the post.

I am curious as to what version of the RiOS you are running?  If you are running version 6, or newer, there is a Health Check available under the Reports > Diagnistics > Health Check.

This can test to see if your LAN and WAN ports are facing their respective network properly.  (see image)

If not, I would advise you upgrade and run the tests.  The RB SH's will run optimization more effectively when connected properly.

0 Kudos

hhmm, I checked my fw-wa-1020 and id-wa-2020 since they are both on ver 6.  The fw-wa-1020 passed, but the id-wa-2020 failed.

0 Kudos

I am also having some trouble configuring these reports.

If I'm reading correctly, the instructions indicate I need to copy the InterfaceID for the interface on the router that's going to the steelhead, then paste that into the OptimizedInterfaceID column for lan0_0 on the steelhead?

Any help is greatly appreciated!

0 Kudos

Question for you guys, have you cross checked the value you've found here from the web reports of you wan accelerator?  As i understand, this only compares pre packets to post packets, but as i understand wan accelerators like riverbed does bi-directional optimization right?  so that means comparing pre and post packet sizes isn't that accurate isn't it?  there has to be a way of polling data for inbound optimization and outbound optimization right?  

0 Kudos

That is exactly what I am finding to be the problem. I'll log into one of our Riverbeds and run an optimization report for the past 24 hours and the data does not match what solarwinds is reporting. That's why I asked before about the setup, to make sure there's nothing I'm missing.

Is this something that simply won't work accurately? I've tried numerous configurations to see if I can get the same results from Solarwinds as the Riverbed, but no matter what I try, the information is different. I'm beginning to come to the conclusion that since optimization does go both ways, that using this method of monitoring (in Solarwinds) for optimization is a no-go.

Can anyone confirm before I waste more time trying to make the software do something it won't?

0 Kudos

BakerD,

I don't know if you are still working on this or have completed now, but the quick answer is yes.  The RB SH's do use the In_Path 0_0 Interfaces when you are set up in In_Path mode.  One of your In_Path 0_0 interfaces faces the WAN and one faces the LAN and these are the two Int's that will be used to configure the WAN Optimization report to calculate data reductions properly.

Int connected to your switch is LAN and Int connected to your router should be WAN

In your RB SH interface go to Reports, Networking, Interface Counters and you will see the below interfaces setup:

In-Path: LAN Interface (inpath0_0 - main interface)

In-Path: WAN Interface (inpath0_0 - main interface)

Make sure your RB SH's are setup properly, else your statistics will be askew.  I came on board after ours were installed and three were connected in reverse.  Now they are correct and data is reflecting proper optimization stats.

Hope this helps.

0 Kudos

I am looking into this too. About half my interfaces show negative numbers in the %Optimized Packets and in % Optimized Bytes. Perhaps i am doing it wrong though.

Should I be taking the InterfaceID of Wan0_0 and putting that in the OptimizedInterfaceID of Lan0_0?

Or should these interface InterfaceID's be the InterfaceID of the port on the switch that the riverbed connects to on the LAN side and the InterfaceID the riverbed connects to on the Router on the WAN side?

0 Kudos

Ok I got my interfaces correct but now I have another issue.

One of my Riverbed devices has both the 0_0 and 0_1 interfaces passing traffic and connected to the same WAN router. Therefore when I add these 2 interfaces to this report each interface only shows part of the traffic on the LAN side but all the traffic on the WAN. This makes the report look like negative optimization.

What I need to do is combine the bytes transferred on both 0_0 and 0_1 and then calculate that against the one common WAN interface.

Is there a way to combine these 2 port together for this calulation?

0 Kudos

I'm still not having any luck.  All our remote offices are setup the same, PCs to switch to lan0_0 on the riverbed then wan0_0 on the riverbed to router.  By following this document "> I get negative %.  The 2nd one in my report below is setup as described in the linked document.  I have tried arranging/assigning different interfaces as you can see in my below picture.  I don't have them hooked up backwards or they wouldn't be working.  The data and graphs provided in the riverbed GUI show positive %.

So is the document wrong?  The screen shot in the document shows positive %.  Maybe the person who created the doc hooked up theirs wrong?

0 Kudos