Performance Benchmarking with TPC-C

On this page Carat arrow pointing down
As of May 12, 2021, CockroachDB v19.2 is no longer supported. For more details, refer to the Release Support Policy.

This page shows you how to reproduce CockroachDB's TPC-C performance benchmarking results on commodity AWS hardware. Across all scales, CockroachDB can process tpmC (new order transactions per minute) at near maximum efficiency. Start by choosing the scale you're interested in:

Warehouses Data size Cluster size
10 2GB 3 nodes on your laptop
1000 80GB 3 nodes on c5d.4xlarge machines
10,000 800GB 15 nodes on c5d.4xlarge machines
100,000 8TB 81 nodes on c5d.9xlarge machines

Before you begin

Step 1. Start CockroachDB

  1. Use the cockroach start command to start 3 nodes:

    $ cockroach start \
    --insecure \
    --store=tpcc-local1 \
    --listen-addr=localhost:26257 \
    --http-addr=localhost:8080 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    $ cockroach start \
    --insecure \
    --store=tpcc-local2 \
    --listen-addr=localhost:26258 \
    --http-addr=localhost:8081 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
    $ cockroach start \
    --insecure \
    --store=tpcc-local3 \
    --listen-addr=localhost:26259 \
    --http-addr=localhost:8082 \
    --join=localhost:26257,localhost:26258,localhost:26259 \
  2. Use the cockroach init command to perform a one-time initialization of the cluster:

    $ cockroach init \
    --insecure \

Step 2. Import the TPC-C dataset

CockroachDB comes with built-in load generators for simulating different types of client workloads, printing out per-operation statistics every second and totals after a specific duration or max number of operations. This step features CockroachDB's version of the TPC-C workload.

Use cockroach workload to load the initial schema and data:

$ cockroach workload init tpcc \
--warehouses=10 \

This will take about ten minutes to load, after which you'll see the following output:

I191024 03:41:34.308865 1 workload/workloadsql/dataload.go:135  imported warehouse (0s, 10 rows)
I191024 03:41:34.353839 1 workload/workloadsql/dataload.go:135  imported district (0s, 100 rows)
I191024 03:42:00.865733 1 workload/workloadsql/dataload.go:135  imported customer (27s, 300000 rows)
I191024 03:42:13.233536 1 workload/workloadsql/dataload.go:135  imported history (12s, 300000 rows)
I191024 03:42:20.893806 1 workload/workloadsql/dataload.go:135  imported order (8s, 300000 rows)
I191024 03:42:21.716409 1 workload/workloadsql/dataload.go:135  imported new_order (1s, 90000 rows)
I191024 03:42:23.483713 1 workload/workloadsql/dataload.go:135  imported item (2s, 100000 rows)
I191024 03:43:37.660918 1 workload/workloadsql/dataload.go:135  imported stock (1m14s, 1000000 rows)
I191024 03:46:51.682670 1 workload/workloadsql/dataload.go:135  imported order_line (3m14s, 3001222 rows)

Step 3. Run the benchmark

Run the workload for ten "warehouses" of data for ten minutes:

$ cockroach workload run tpcc \
--warehouses=10 \
--ramp=3m \
--duration=10m \

You'll see per-operation statistics every second:

Initializing 20 connections...
Initializing 100 workers and preparing statements...
    1.0s        0            0.0            0.0      0.0      0.0      0.0      0.0 delivery
    1.0s        0            0.0            0.0      0.0      0.0      0.0      0.0 newOrder
  105.0s        0            0.0            0.2      0.0      0.0      0.0      0.0 delivery
  105.0s        0            4.0            1.8     44.0     46.1     46.1     46.1 newOrder
  105.0s        0            0.0            0.2      0.0      0.0      0.0      0.0 orderStatus
  105.0s        0            1.0            2.0     14.7     14.7     14.7     14.7 payment
  105.0s        0            0.0            0.2      0.0      0.0      0.0      0.0 stockLevel

For more tpcc options, use cockroach workload run tpcc --help. For details about other built-in load generators, use cockroach workload run --help.

Step 4. Interpret the results

Once the workload has finished running, you'll see a final output line:

  300.0s      121.6  94.6%     41.0     39.8     54.5     71.3     96.5    130.0

You will also see some audit checks and latency statistics for each individual query. For this run, some of those checks might indicate that they were SKIPPED due to insufficient data. For a more comprehensive test, run workload for a longer duration (e.g., two hours). The tpmC (new order transactions/minute) number is the headline number and efc ("efficiency") tells you how close CockroachDB gets to theoretical maximum tpmC.

The TPC-C specification has p90 latency requirements in the order of seconds, but as you see here, CockroachDB far surpasses that requirement with p90 latencies in the tens of milliseconds.

Step 5. Clean up

  1. When you're done with your test cluster, use the cockroach quit command to gracefully shut down each node.

    $ cockroach quit --insecure --host=localhost:26257
    $ cockroach quit --insecure --host=localhost:26258

    For the last node, the shutdown process will take longer (about a minute each) and will eventually force the node to stop. This is because, with only 1 of 3 nodes left, all ranges no longer have a majority of replicas available, and so the cluster is no longer operational.

    $ cockroach quit --insecure --host=localhost:26259
  2. To restart the cluster at a later time, run the same cockroach start commands as earlier from the directory containing the nodes' data stores.

    If you do not plan to restart the cluster, you may want to remove the nodes' data stores:

    $ rm -rf tpcc-local1 tpcc-local2 tpcc-local3

Yes No
On this page

Yes No