The Daily Network Monitor

Browsing around, I found this piece in India New England (yes, that’s the name of the publication) called Network downtime can be big expense for business.  Tim Hebert is definitely singing our song.  To wit:

Unplanned downtime is what keeps IT professionals, executives and business owners awake at night. Natural disasters and utility failures only account for three percent of all outages. Hardware failures account for less than 10 percent of all network failures. Systems errors account for less than eight percent of failures; application errors, 19 percent.

Industry experts estimate that almost 60 percent of network failures are caused by human error. This problem can be attacked through better training for IT organizations and end-users, better network documentation, better change-management controls and processes and better network monitoring and management.

Sixty percent of network failures caused by human error?  Wow.  There’s no citation to back that up, but it has the ring of truth.  After all, if downtime came from more controllable sources, there wouldn’t be so much of it, right?

In any case, Tim’s piece makes clear the case for effective network management and network monitoring.  I hope IT managers take heed and make a small investment in protection against what could be a huge loss.

I was wandering around various job sites just to see what kinds of network management career opportunities are available. I was looking for job requirements where SLAs are defined as well as what level of experience was required. And while I did find a lot of opportunities where SLAs were defined, another interesting aspect became apparent, which I discuss below.

While the normal requirements for Cisco certifications, VoIP, QoS and routing, server monitoring experience were required, one thing that stood out was the need for experience with network monitoring tools. Of course the big-boys, HP, IBM and CA were mentioned but even more surprising to me was how many positions defined a requirement for experience with WhatsUp Gold.

While I am an Ipswitch employee, it was very interesting to see the number of positions that required WhatsUp experience as a requirement. Below is one job description that was particularly interesting (names have been deleted to protect the innocent):

Dynamic company is currently seeking Telecom VOIP Administrator. Candidate will assist in managing WAN connectivity to 50+ remote sites; participate in 24×7 on-call rotation; manage vendor relations; provide 2nd tier support to helpdesk staff; assist in managing existing Cisco IP Telephony solution and installing solution at new sites; document WAN and LAN, keeping documentation up-to-date.

Technical Experience:
Windows XP, Windows Server 2003, Microsoft Office, Microsoft Visio, Active Directory, Cisco Networking, EIGRP, BGP, Frame-relay, MPLS, Network Monitoring, WhatsUp Gold, SonicWALL Firewalls, Catalyst 45xx, Cisco IP Telephony, Cisco Call Manager 4.0, Cisco Unity voicemail

Requirements:
CCNA, College Degree (B.S. or B.A.), 3+ years Telecom experience, Cisco VoIP experience

This leads me to ask a question of network administrators/managers out there. Would a certification on WhatsUp Gold be relevant to building your career credentials?

In an earlier entry, Back into the Fray, I listed what has changed and has not changed after I left enterprise networking and I joined Ipswitch. One of the items that changed was VoIP. VoIP seems to have fallen under what is now termed unified communications.

Both Microsoft and Cisco have staked places at the unified communications table. But what does unified communications really mean. Is it VoIP? Is it IM? Is it collaboration? Is it email? Or is it all of these things melded into one?

What ever it is, it means only one thing to network managers. How much effort is it going to take to manage?

From this one question we can deduce a number of other implications to an already saturated infrastructure and the ability to manage yet another cool technology someone just had to have.

If it is server centric, read Microsoft, this means more server focused hardware to manage. How will this server based infrastructure be managed? Not only is there additional server hardware to manage, but also license management (read CALs) to ensure EULA compliance. 4000 IP phones, means 4000 CALs, unless Microsoft is changing their licensing model.

Or networking gear centric, read Cisco, this fits nicely with most existing installed infrastructures and most of the management capabilities are already in place.

QoS management for VoIP is key to the whole effort of unified communications, QoS is network centric not server centric.

I’m not trying to take a slanted view of one company over another, just what make sense for an organization. If it was a network that I was responsible for, I would choose the network centric approach over the application centric approach every time.