3 ms·
Their reliability only looks bad in comparison with other cloud providers. Microsoft can still probably do a much better job compared to a small team.
by outworlder 6y ago
Their reliability only looks bad in comparison with other cloud providers.
Microsoft can still probably do a much better job compared to a small team.
- Nextgrid 6y agoThe variable here is not just the size & skill of the team but their needs. In most cases a single company or team will need a simpler, smaller-scale deployment, probably running on one or a handful of machines and not requiring the insane complexity that a public cloud provider is running behind the scenes to provide their service. This eliminates a lot of moving parts and things that can fail. Imagine an Nginx server running on a single machine. There's very little that can go wrong here beyond hardware failure (and certain types of failures can be mitigated with things like RAID), and yet it is probably enough to host most internal websites. Now compare this to something like Azure App Service which is obviously more complex to be able to support many tenants, load-balance, etc. There is much more that can go wrong with the entire App Service infrastructure (due to its complexity and moving parts) than with a single machine running Nginx, and the complexity will also delay disaster recovery efforts (this outage is now lasting for more than an hour - you can reinstall an entire Linux web server from scratch in half that time).
- userbinator 6y agoAn internal team is also going to be far more accessible and responsive to the company's needs.
- acwan93 6y agoIf they have enough resources. When everyone is busy working on the revenue generating products for the company, taking time away to work on internal network infrastructure immediately becomes a cost center.