09-15-2013 06:47 PM - edited 03-04-2019 09:02 PM
Hi there,
I currently have two locations connected using a MetroE link of 50Mpbs (CIR) burstable up to a 100Mbps. Location A has all our servers and storage and in location B, all the important departments such as finance, etc. I use Solarwinds to monitor the MetroE router interfaces but I don't see a high utilization even though I see there is a high volume of traffic between my core and Location A's MetroE router Lan interface. The next thing I'll be doing is checking the interface itself in the router by sampling the traffic in a 30s interval. Any other suggestions on this?
The other question I have is that I am doing L3 on those router with EIGRP and I am in the process of implementing QoS but I wanted to know whether or not I need to match the carrier's marking at the L2 level. If so, what questions should I ask them on this regard?
Thanks in advance.
Rafael.
09-15-2013 07:36 PM
Regarding the link monitoring, it could be an snmp or the software, try different software or check if there is another mib, also you can enable netflow which will give you more details about the utilization if he link
For qos, check if your ISP is doing any qos remarking, most probably they do. Ask them which one to use either cos or dscp, they can provide you this answer based on their network design
Sent from Cisco Technical Support iPhone App
09-16-2013 01:52 AM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
The other question I have is that I am doing L3 on those router with EIGRP and I am in the process of implementing QoS but I wanted to know whether or not I need to match the carrier's marking at the L2 level. If so, what questions should I ask them on this regard?
If the ports on both sides only have to deal with traffic from just the one other side, your QoS needs should be met by your router QoS features.
If the ports have multiple far side partners, then QoS goals can be very difficult to accomplish unless you can leverage provider QoS. If you do work with provider QoS, the first question is what QoS can they provide. Often providers have a limited set of QoS features they support. Sometimes they charge more for using them, or some of them.
09-18-2013 09:44 AM
Thank you for your reply.
I want to hear your opinions about having all servers and storage at one location while half of your users sit at another location connected with a 100Mbps. There are only so many queues that you can created in QoS; therefore, not all applications can be prioritized.
Tks, Rafael.
09-18-2013 10:24 AM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
There are only so many queues that you can created in QoS; therefore, not all applications can be prioritized.
Actually if your try to prioritize all applications, effectively there's no prioritization.
As to number of queues, some platforms can support quite a few flow queues, but even when that's not possible you can obtain much benefit from 2 to 4 queues, if you use them intelligently and if there's sufficient bandwidth for the traffic that needs a certain amount of bandwidth. Some traffic (by design) will try to use all available bandwidth even when less bandwidth would satisfy the application. It's this kind of traffic, mixed with traffic that has stringent bandwidth requirements, where QoS can make a huge difference.
I want to hear your opinions about having all servers and storage at one location while half of your users sit at another location connected with a 100Mbps.
Might be perfectly fine or totally horrible; lots of variables. Even when you have lots of available bandwidth, end-to-end latency can be very detimental to "performance" of some applications.
09-19-2013 02:47 PM
I just want to understand how companies can keep their data centers at other locations and no have delay/slowness issues. I would say that a good 75% to 80% traffic is associated with file shared access which something I believe doesn't need any type of prioritization. Now, users complaint about how slow their access is and when I look at the link there is not a lot of traffic going on.
I use Solarwinds to monitor links as well as the router's command line. If anyone can explain how to accomplish this or where to look, it would be greatly appreciated.
09-19-2013 02:53 PM
When I had to worry about this stuff in a previous position I found that Link bandwidth was rarely the issue.
tuning the TCP stack can do wonders, along with QOS.
http://www.ipbalance.com/traffic-analysis/throughput/104-tcp-throughput-calculation-formula.html
09-19-2013 05:05 PM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
I just want to understand how companies can keep their data centers at other locations and no have delay/slowness issues.
Who says they don't have issues?
You can, if you can afford it, but bandwidth. Distance based latency is a more difficult to deal with. WAAS appliances can help (sometimes impressively so), but it's never quite the same as "being there" (i.e. local LAN latency).
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide