3 ms·
Ah, so your systems are down until you fix a node, and if the drive(s) fail you have a dead node you're stuck restoring backups. I guess that's fine, I'm just
by dijit 5y ago
Ah, so your systems are down until you fix a node, and if the drive(s) fail you have a dead node you're stuck restoring backups.
I guess that's fine, I'm just used to having labs to test how I can avoid single instance failures. :)
- Axsuul 5y agoThe path to redundancy is there if you need it but it's not something you have to go with from the start. You can also run a database with replicas to handle any drive failures, etc. I have been running a production-grade system with thousands of customers like this for a few years.