INTEROP 2005

By Dipendra Bantawa

Interop 2005, the infrastructure trade show, came to New York in the fall. I will not cover each and every exhibitor, or every featured presentation. Instead I will focus on some products exhibited at the show that have good potential for being integrated with current Systems Management products.

Avaya ExpertNet VoIP Assessment Tool is an example of a product for extending Systems Management deployments. Other vendors, such as NetIQ, also have VoIP assessment tools assess the readiness of a network to support voice and video traffic. The reason I liked the product is that there is huge need for pre-SM deployment assessment.

NetQoS uses what it calls SuperAgent to monitor end-to-end performance passively without deploying any desktop or server agents. SuperAgent separates delays due to applications, network and server, enabling more rapid troubleshooting. This product almost acts like “SI”, the performance monitoring product from Netuitive. It generates baselines and compares the metrics that it collects and automatically tries to investigate the cause of the problem, as it gathers data, after it detects a problem by capturing filtered packet data, SNMP polling and traceroutes.

Network Instruments nTAPs provide monitoring devices with access to all network traffic without disrupting or adding any more network traffic. In simple terms, it makes a copy of the traffic that flows in and out that can be fed to any analysis tool or monitoring system. This out of band monitoring approach can be helpful when the network/infrastructure is already severely suffering as the monitoring system will only add more traffic and use already over utilized resources.

We at JMACINC are always interested in serious forums such as Interop that offer a broad range of innovations for optimizing and extending the leading Systems Management products.

Refer to www.JMACINC.com for the rest of the story.

JFFNMS: Another Half-baked Idea

I couldn't help noticing the offer for a "FREE Monitoring Tool Inside" on the cover of this week's issue of "Windows IT Pro" magazine, a Penton publication. Unfortunately, a cursory glance at the contents showed me that this FREE solution was more of a curiosity than a really useful tool.

Why a "curiosity"? Because it is assembled from a number of open systems tools, and that's a good thing.

What would make it a "really useful tool"? The ability to function as an agent on the server on a 24x7 basis. What this tool offers, as do so many other partial solutions, is a central poller facility rather than a locally deployed agent. I know many IT shops that have developed their own monitoring tools to run as an agent service simply by conversion of their programs into a Windows service. Without the ability to run on the local agent, you can't alert when there are network problems, and you can't remediate the problems until the network is back up.

If you're a small shop seriously looking to implement a monitoring solution with as few pesos as possible, consider JFFNMS only as a curiosity.

Refer to www.JMACINC.com for the rest of the story.

Updates to AppManager 6.0 Review

Last month I posted a review of the beta version of AppManager 6.0. This month (July 15, 2004) I assisted NetIQ by demonstrating some of the new features of the product in a Placeware audiocast conducted by NetIQ.

There are some features of the product that were discussed in the audiocast but were not covered in the review, and they include the following:

1. Action_RunKS -- this knowledge script allows the administrator to launch up to three knowledge scripts dynamically from a job action
2. Ability to resize the values tab of a knowledge script
3. Knowledge scripts for collecting data for the Diagnostic Console. This integration with Diatnostic Console includes both Exchange and NT server core metrics.
4. Action Severity Configuration -- this refers to the ability within many Action knowledge scripts to fire only when the severity of the triggering event is within a range defined as part of the Action knowledge script.

As I think of other features I've left out, I'll post them here.

2003 Market Research on SLAs from Veritas

In a September 2003 survey commissioned by Veritas, data-center managers and their non-IT counterparts at 604 organizations with at least 500 employees responded to questions on the nature of their usage of SLAs. Veritas sponsored the survey as part of its "Utility Computing" product focus.

I've culled some of their statistics here, but you can download the original report for the full analysis.

1. Over 59% of the respondents do use SLAs.

2. Among organizations that have SLAs in place, the following key IT performance areas are covered:

- Processing performance (37 percent)
- System availability and uptime (35 percent)
- Restoration times following an outage (29 percent)
- None of the above (39 percent)

3. Twenty five percent of all cases said the SLAs were crafted without the involvement of the department heads.

4. With respect to the IT reports generated in support of SLAs, non-IT managers used them as follows:

- to make department operations more efficient (34 percent)
- to work with the IT department to lower costs (28 percent)
- not used successfully (39 percent)

The research covered a wide array of companies in the United States, nine countries throughout Europe, the Middle East and South Africa. It was conducted by Dynamic Markets of the UK.