When it comes to correctness guarantees, there are few communities in Computer Science (software) more obsessed with them than the database and distributed system communities. They better be serious about correctness because everyone's bank account balances depend on it. You have heard about terms like Paxos, consensus, fault tolerance, but…
std::enable_shared_from_this
[https://en.cppreference.com/w/cpp/memory/enable_shared_from_this] allows a
class to have a valid shared_ptr of this. Simply adding a member function that
returns shared_ptr
FoundationDB is a very impressive database. Its paper won the best industry paper award in SIGMOD'21. In this post, I will explain, in detail, how FDB works and discuss a few very interesting design choices they made. It's a dense paper packed with neat ideas. Many details (sometimes even proof…
Paxos, is the famous synonym for consistency in the context of distributed system. Unfortunately, consistency alone, is such an overloaded term, it's often practically meaningless. In this post, I will explain the difference between Paxos-consistency vs. quorum-consistency. I assume you know what Paxos is and what problem it solves. If…
You have heard about Coroutine in C++, and you want to use it. There're two coroutine implementations that are considered most mature - cppcoro and folly::coro. They are written by the same guy - Lewis Baker. He's brilliant, and you shoud watch his cppcon talk on structured concurrency [https:…
struct Foo {
// virtual ~Foo() {};
int a;
};
struct Bar : public Foo {
~Bar() {std::cout << "bar dtor" << std::endl;};
};
int main() {
std::shared_ptr
You have created tables with AUTO_INCREMENT. Something like CREATE TABLE user ( user_id INT(11) NOT NULL AUTO_INCREMENT, ... ) MySQL manages user_id for you automatically. You get user_id 1, 2, 3, 4, ... But how does it work? InnoDB, one of the most common storage engine MySQL uses,…
Sequential lock is a common technique used to protect data that's frequently read and rarely updated. Writers are required to take exclusive lock to mutate the data. Reads are effectively lock free and optimistic. The data protected by a SeqLock usually needs to be trivially_copy_constructible [https://en.cppreference.…