We need someone to look after ops and site reliability stuff.
This job is in Bristol (UK) with our Delib business. We provide web apps for government customers around the world. Citizens use them to participate in decision making. We need reliable production environments for them.
We have a decent environment (in Bristol) where we write software. It’s not perfect; it’s a bit busy, but we care and say thanks, and the people are ok and I reckon that counts for a lot. We also have version control and testing and decent chairs and lunch, a lack of fear and no pissy politics.
Typically each customer has their own app instance. Lots of the stuff we do is sensitive formal policy consultation. It’s important to the people running it and to the people taking part, so we do look for lots of uptime and we never want our customers to end up in judicial review on a policy consultation due to an ops screw up by us.
We also handle sensitive personal data, and we have no intention of screwing that up either.
We aren’t looking for someone to just work down a task list that already exists. We have some ideas about the kind of configurations we like but no consensus. The first thing you’ll be doing is gathering requirements, investigating alternatives and coming up with a proposal to modernise our hosting environment.
We need someone who will work with developers directly, no “sysadmin in the basement”.
- Need to support the development environment as well as production. Don’t need to do the office laptops, printers etc, we have that covered.
- We use Pingdom, FreeBSD and some flavours of Linux.
- We have ticketing, and version control and all the usual stuff.
- We’re not using ITIL or ISO 27001, but if you have experience with those we’d be interested in hearing about it.
- We write our own software in Python, so if you have to write tools then Python is preferred. We also (mostly) like Bourne shell.
- We prefer rational choices to personal preferences.
- Servers are around the world because our customers really need data to stay physically in their territory.
Job involves on-call. Currently developers get woken up by Pingdom, it pisses them off and they moan at me. We pay some kind of get-out-of-bed fee, although ideally you automate all this nonsense away. Plan is to put our next ops hire on an Australian timezone, dunno when that will be though.
Plenty to do.
It’s probably a full-time permanent position, but we’ll consider flexible ways of getting stuff done.
Interested? Drop Tom a line: firstname.lastname@example.org