nv-l
[Top] [All Lists]

Re: Service Level & Bandwidth Monitoring

To: nv-l@lists.tivoli.com
Subject: Re: Service Level & Bandwidth Monitoring
From: davids@ibk.com.au
Date: Tue, 6 Feb 2001 18:16:18 +1100
Danny,

There has been some good feedback on point 1) and the only thing I would
add is that on a busy network I have seen packet loss that resulted in
regular Node and Interface Down messages from Netmon. We had to wind out
the timeout and retry parameters to ensure the outages were genuine.

Regarding point 2) I have used the thresholding capability within Data
Collection to fire a trap to alert me to extended bursts of traffic
associated with large file transfers. This only allows you to isolate the
occurrences to a particular link, we then used RMON probes on the major
LANs to isolate the devices. This sleuth work has to be done at the time of
the occurrence as the averaging effect within RMON can make it difficult to
isolate the culprit after the event.

None of this was terribly satisfactory, so the architectural answer was to
implement Protocol Prioritisation in the routers and put FTP, SMTP, HTTP
traffic etc in a low priority queue so that it didn't effect the
time-critical business applications. Life was much easier after that.

David Steed
Enterprise Management Consultant
Innovative Business Knowledge Pty Ltd
Phone 61-2-9584 8771 Mobile 0402 073 569
Web Page: www.ibk.com.au


> -----Original Message-----
> From:         danny@uk.ibm.com [SMTP:danny@uk.ibm.com]
> Sent:         Monday, 05 February, 2001 09:37
> To:           nv-l@tkg.com
> Subject:           [NV-L] Service Level & Bandwidth Monitoring
>
>
>
> Hi All
>
> These are architectural questions rather than a techinical ones.
>
> 1. Have any of you have used NetView in anger as a tool to monitor the
> service levels being provided by a network provider? My initial thoughts
> are that one could correlate interface down & up traps/events to
establish
> the amount of time a line was unavailable and then put this informaton in
> to a file for offline processing.
>
> 2. I know that NetView is capable of monitoring the throughput on network
> interfaces, but have any of you used NetView to pinpoint large data
> transfers and the associated bottlenecks?
>
> Any information that you guys (and gals) can provide on these items would
> be greatly appreciated.
>
> It may be that there are more appropriate tools for each of these
> functions
> - if so, please net me know.
>
> Thanks in advance for your assistance.
>
> Danny Williams
> Tivoli & AIX Specialist
> IBM Global Services - Integrated Technology Services
> mailto:danny@uk.ibm.com
> Tel: 665520 (01926-465520) -- MOBX: 275520 (07967-275520)
>
>


<Prev in Thread] Current Thread [Next in Thread>

Archive operated by Skills 1st Ltd

See also: The NetView Web