Single-node durable deployment
Run Krishiv as a local daemon with RocksDB state and local shuffle.
The single-node daemon gives you restart-durable state and shuffle on a single host. It is the right starting point for production-style deployments that do not yet need a cluster.
Build
# GCC 15 hosts need this workaround for rocksdb
CXXFLAGS="-include cstdint" cargo build -p krishiv --features single-node --release
Start the daemon
./target/release/krishiv server start --coordinator-addr 0.0.0.0:50051 --durability single-node-durable --checkpoint-dir /var/krishiv/checkpoints
Connect from a client
export KRISHIV_COORDINATOR=http://localhost:50051
import krishiv as ks
session = ks.Session.from_env() # reads KRISHIV_COORDINATOR
use krishiv_api::Session;
let session = Session::from_env().await?;
What each durability profile gives you
| Profile | State | Shuffle | Checkpoints |
|---|---|---|---|
dev-local | In-memory | In-memory | Ephemeral |
single-node-durable | RocksDB | Local disk | Local filesystem |
distributed-durable | RocksDB | Tiered (local + object store) | Object store + etcd |
See the Deployment and Execution Model pages for the full picture.