![]() ![]() The scalability of a system is its ability to cope with increased load. Bonus: this helps with maintainability, e.g no need for planned downtimes. Hardware redundancy is not sustainable with data volumes and/or computing demand. ![]() Prefer software fault-tolerance over hardware redundancy.Generally less correlation in hardware faults than in software bugs: example of Linux’ kernel bug with the leap second of June 30, 2012.security: there’s no tolerating users’ data being compromised Prefer tolerating faults (easier) over preventing faults.Hardware faults (MTTF) / Software faults / Human errors.Netflix Chaos Monkey, then Simian Army: an example of chaos engineering.Distinction between fault (more a less a single component not working as expected) and failure (refers to the whole system).It can refer to noSQL, non-ACID semi-structured data applications like Redis. A datastore is an ill-defined term that design a data system in between data warehouses and data lakes. Part 1: Foundations of Data Systems Chapter 1: Reliable, scalable and maintainable applications ![]()
0 Comments
Leave a Reply. |