Things over at $CURRENT_EMPLOYER have been pretty boring lately.
I mean, I had to explain to a “client” (that's “inside the company customer”) over the past few days that yes, it really would cause a problem for the backup infrastructure if their systems, configured as media servers (meaning, polled, via TCP/IP, every ten minutes by a daemon on the master server) were just randomly off the network while being on the other side of a firewall, meaning that the daemon just hung for TCP/IP-handshake-timeout every time it tried, pre-empting other slightly more important tasks, like doing backups for systems with the courtesy to be, you know, on the network. And then, today, that no, we really wouldn't be able to connect to their addresses, one in 155/8 and the other in 10/8, if they were bound to the same ethernet interface. But that's all kids' stuff.
I guess I may have a dork post in me about how NetBackup doesn't scale so good. (Support asks for “VERBOSE = 5 logs”. I point out that that will run us out of–24 GB of–log disk in about 4 hours, and would they maybe like to be more specific about which daemons they'd like verbose logs for. Or they ask for the support script to be run on a media server that can't reach much besides the master because of a firewall, and I point out that it'll take about 13 hours for that to run, because of the firewall, and the support script's insipid need to poll every other media server for each of the 20000+ volumes in the volume DB… for bonus points: “How many media servers do you have?” “Oh, something around 50, depending on who else did how many SSO configs this week.” “buhbuhbubh” “Could you escalate this call, please?”) But, really, it's getting to be like shooting magnetic fish in a champagne flute with lead bullets.
Maybe I should dust off the resume or something. Since, clearly, the purpose of my employment is to generate immediately frustrating but eventually amusing anecdotes for forums such as this.
Post a Comment