I was recently hired for a lateral move within OmniMegaCorp to another department with slightly different duties. I’m really excited about this move, because I get to get more involved with an area where I think computing is going: grid or “cluster” virtualized computing.
I was recently hired for a lateral move within OmniMegaCorp to another department with slightly different duties. I’m really excited about this move, because I get to get more involved with an area where I think computing is going: grid or “cluster” virtualized computing. These fault-tolerant, highly-reliable systems bring some amazing benefits for computing in environments that don’t demand the very highest performance from bare-metal machines. That said, they still offer substantial performance for very large infrastructure projects that tolerate latency well, and the systems-management and uptime benefits alone are enough for many to be willing to sacrifice a little performance for immense reliability.
As part of my transfer — which will involve sitting at the same desk I’ve been at for nearly half a decade — I was asked to write a job description of what I do in order to aid my current management in finding a replacement. I thought it might be entertaining for you, too.
Duck Soup Wrote: > 1. Projects that you are currently working on or soon to be involved with > 2. Functions that you are the primary contact for > 3. Days and duty at our failover site
To be honest, other that deployments, I have very few “projects” since so much of my work is standard duties and firefighting. That’s one of the things I look forward to in this move: more time to oversee projects from beginning to end, rather than being so reactive all the time.
Projects: * Linux/Solaris deployment wave 2. Given that our mass-deployment product doesn’t seem to be working on these subnets for Linux systems, this can be very on-site time-consuming. Check w/E. Honda for exact times. Main jobs are configuring the DRACs, ensuring network port requests go through, DNS coordination, making sure power/network is set up, assigning/labeling machines, installing base OS. * Ongoing migration of various Phase 1 and legacy machines to Xen/VMWare and alternative operating systems.
Standard duties: * Failover payment site backup rotation and maintenance (~2-3 hours per week), usually Thursday mornings. Roger Giggley and I shared this duty before, and they are production systems. I volunteered to do it each Thursday because it’s on my way to work, and quite a bit out of his way. * Backup monitoring, maintenance, and rotations: 10+ hours/week. About 3-5 hours actually rotating tapes. * Restores. This requires an on-site person to swap in tapes much of the time. 2-10 hours per week (depends where we are in the build cycle). * One-time backups. Often requires short-term ejects. Averages just a few hours per week, around 2-3 hours. * Service request queue. Rest of the week, performing miscellaneous on-site Linux/Solaris/HP-UX/AIX troubleshooting and maintenance. (~20 hours/week) * Attend Friday morning coordination meeting @ 10:00 AM US/Mountain. Provide reports and status updates for coordination with Logistics, IT, and Space/Power. Also report on downtime, if any. (~1 hour/week) * Provide support on general Linux/UNIX questions to peers/managers/directors in other development areas (~1 hour/week) * Provide on-site support for remote members of my team and other UNIX, storage, and backup teams who have encountered difficulty administering systems remotely (up to 10 hours/week) * Provide on-site support for other teams that require UNIX expertise, though not necessarily intervention (2-3 hours/week). * Handle auditing duties gracefully with external auditors. (a couple times a year) * Interface with vendors for hardware replacements and repair. (mostly OtherBigVendorCompetitor for backup systems and NAS)
Duties recently removed: * NAS administration (went to storage team) * SAN administration (went to storage team)
Useful skills for replacement: * Thorough understanding of Netbackup, robotic tape libraries, and fibre-channel architecture. * Linux, AIX, HP-UX, and Solaris administration abilities (though maybe the Linux stuff could be shuffled off to the new Linux team) * Demonstrate ability to automate processes with at least one scripting language (Perl and Posix Shell or Bash a plus) * Show familiarity with kernel rebuilding, performance tuning, and basic performance troubleshooting (familiar with “top”, “free”, “iostat”, “find”, etc.) * Be able to write an init script from scratch, including case() statements to handle at least stop/start requirement on Linux/Solaris. * Be familiar with at least one UNIX rapid deployment system (AIX: NIM, Solaris: Jumpstart, Linux: Kickstart, HP-UX: Ignite) * Familiarity with Xen, VMWare, and other virtualization technologies. * Self-directed with ability to work under severe time pressure and while juggling many equally-important high-priority items. * Able to communicate effectively via electronic mail, instant messenger, and phone. * Must be willing to carry mobile phone 24/7 and be available for on-call duties on occasional weekends. * Must be able to lift seventy pounds. * Willing to work strange hours occasionally to handle joint issues with staff in India.
So what does your job look like?