- First set some variables.
export aws_region="eu-west-1"
export aws_vpc_cidr="10.2.0.0/16"
export clus2="mb-tpcc-cluster-1"
export kubernetes_version=1.21
- Create you eks cluster of the desired size here I am using three instances of type
c5d.4xlarge.
eksctl create cluster \
--name $clus2 \
--nodegroup-name standard-workers \
--node-type c5d.4xlarge \
--nodes 3 \
--region $aws_region \
--vpc-cidr $aws_vpc_cidr \
--version $kubernetes_version
- Download the stateful set configuration and edit to meet your requirements.
curl -O https://raw.githubusercontent.com/cockroachdb/cockroach/master/cloud/kubernetes/bring-your-own-certs/cockroachdb-statefulset.yaml
The key element to update in the config file is the resources section. You will want to edit these to make the most of the instances in your worker pool. In my example, my instances have 16 vCPUs and 32GiB RAM. You need to allow some resources for your OS and Kubernetes components.
Example Output:
resources:
requests:
cpu: "12"
memory: "24Gi"
limits:
cpu: "12"
memory: "24Gi"
- Make these directories for our certificates.
mkdir certs my-safe-directory
- Use the cockroach CLI to create the ca.
cockroach cert create-ca --certs-dir=certs --ca-key=my-safe-directory/ca.key
- Create a client certificate and key pair for the root user.
cockroach cert create-client \
root \
--certs-dir=certs \
--ca-key=my-safe-directory/ca.key
- Upload the client certificate and key to the Kubernetes cluster as a secret.
kubectl create secret \
generic cockroachdb.client.root \
--from-file=certs
- Create the certificate and key pair for your CockroachDB nodes.
cockroach cert create-node \
localhost 127.0.0.1 \
cockroachdb-public \
cockroachdb-public.default \
cockroachdb-public.default.svc.cluster.local \
'*.cockroachdb' \
'*.cockroachdb.default' \
'*.cockroachdb.default.svc.cluster.local' \
--certs-dir=certs \
--ca-key=my-safe-directory/ca.key
- Upload the node certificate and key to the Kubernetes cluster as a secret.
kubectl create secret \
generic cockroachdb.node \
--from-file=certs
- Create the kubernetes resources.
kubectl create -f cockroachdb-statefulset.yaml
kubectl create -f aws-svc-admin-ui.yaml
- Run cockroach init on one of the pods to complete the node startup process and have them join together as a cluster.
kubectl exec -it cockroachdb-0 \
-- /cockroach/cockroach init \
--certs-dir=/cockroach/cockroach-certs
- To use the CockroachDB SQL client, first launch a secure pod running the cockroach binary.
kubectl create -f https://raw.githubusercontent.com/cockroachdb/cockroach/master/cloud/kubernetes/bring-your-own-certs/client.yaml
kubectl exec -it cockroachdb-client-secure \
-- ./cockroach sql \
--certs-dir=/cockroach-certs \
--host=cockroachdb-public
- Create a user to log in to the UI with.
CREATE USER craig WITH PASSWORD 'cockroach';
GRANT admin TO craig;
\q
- Now shell into the secure client again but this time just with
shas we need to pull down our dataset form the internet.
kubectl exec -it cockroachdb-client-secure -- sh
- Download the CockroachDB archive for Linux, extract the binary, and copy it into the PATH.
curl https://binaries.cockroachdb.com/cockroach-v22.1.1.linux-amd64.tgz | tar -xz
cp -i cockroach-v22.1.1.linux-amd64/cockroach /usr/local/bin/
If you get a permissions error, prefix the command with sudo.
- Import the TPC-C dataset.
cockroach workload fixtures import tpcc --warehouses=2500 'postgresql://root@cockroachdb-0.cockroachdb:26257?sslmode=verify-full&sslrootcert=/cockroach-certs/ca.crt&sslcert=/cockroach-certs/client.root.crt&sslkey=/cockroach-certs/client.root.key'
Example Output:
I220616 09:49:47.527128 1 ccl/workloadccl/fixture.go:318 [-] 1 starting import of 9 tables
I220616 09:49:49.874109 42 ccl/workloadccl/fixture.go:483 [-] 2 imported 2.5 MiB in district table (25000 rows, 0 index entries, took 2.27973529s, 1.08 MiB/s)
I220616 09:49:50.389726 41 ccl/workloadccl/fixture.go:483 [-] 3 imported 133 KiB in warehouse table (2500 rows, 0 index entries, took 2.795448522s, 0.05 MiB/s)
I220616 09:49:50.490800 47 ccl/workloadccl/fixture.go:483 [-] 4 imported 7.9 MiB in item table (100000 rows, 0 index entries, took 2.896393934s, 2.72 MiB/s)
I220616 09:50:15.562040 46 ccl/workloadccl/fixture.go:483 [-] 5 imported 340 MiB in new_order table (22500000 rows, 0 index entries, took 27.967682747s, 12.16 MiB/s)
I220616 09:52:52.223157 45 ccl/workloadccl/fixture.go:483 [-] 6 imported 4.1 GiB in order table (75000000 rows, 75000000 index entries, took 3m4.628814502s, 22.55 MiB/s)
I220616 09:53:10.150226 44 ccl/workloadccl/fixture.go:483 [-] 7 imported 5.4 GiB in history table (75000000 rows, 0 index entries, took 3m22.555905007s, 27.20 MiB/s)
I220616 09:59:07.335145 43 ccl/workloadccl/fixture.go:483 [-] 8 imported 43 GiB in customer table (75000000 rows, 75000000 index entries, took 9m19.740752911s, 78.92 MiB/s)
I220616 10:02:36.886581 49 ccl/workloadccl/fixture.go:483 [-] 9 imported 43 GiB in order_line table (750022630 rows, 0 index entries, took 12m49.292182104s, 57.13 MiB/s)
I220616 10:02:40.529454 48 ccl/workloadccl/fixture.go:483 [-] 10 imported 75 GiB in stock table (250000000 rows, 0 index entries, took 12m52.935031273s, 99.82 MiB/s)
I220616 10:02:40.554137 1 ccl/workloadccl/fixture.go:326 [-] 11 imported 171 GiB bytes in 9 tables (took 12m53.026841034s, 226.76 MiB/s)
I220616 10:02:41.646193 1 ccl/workloadccl/cliccl/fixtures.go:343 [-] 12 fixture is restored; now running consistency checks (ctrl-c to abort)
I220616 10:02:41.682992 1 workload/tpcc/tpcc.go:517 [-] 13 check 3.3.2.1 took 36.744891ms
I220616 10:02:59.700697 1 workload/tpcc/tpcc.go:517 [-] 14 check 3.3.2.2 took 18.017641828s
I220616 10:03:02.729988 1 workload/tpcc/tpcc.go:517 [-] 15 check 3.3.2.3 took 3.02924782s
I220616 10:05:39.619676 1 workload/tpcc/tpcc.go:517 [-] 16 check 3.3.2.4 took 2m36.889645739s
I220616 10:06:42.635846 1 workload/tpcc/tpcc.go:517 [-] 17 check 3.3.2.5 took 1m3.016103174s
I220616 10:10:47.545607 1 workload/tpcc/tpcc.go:517 [-] 18 check 3.3.2.7 took 4m4.909679287s
I220616 10:11:11.850450 1 workload/tpcc/tpcc.go:517 [-] 19 check 3.3.2.8 took 24.304779361s
I220616 10:11:42.313118 1 workload/tpcc/tpcc.go:517 [-] 20 check 3.3.2.9 took 30.46262877s
- Create the a file called
addrson the file system with the command below.
cat > addrs<<EOF
postgresql://root@cockroachdb-0.cockroachdb:26257?sslmode=verify-full&sslrootcert=/cockroach-certs/ca.crt&sslcert=/cockroach-certs/client.root.crt&sslkey=/cockroach-certs/client.root.key postgresql://root@cockroachdb-1.cockroachdb:26257?sslmode=verify-full&sslrootcert=/cockroach-certs/ca.crt&sslcert=/cockroach-certs/client.root.crt&sslkey=/cockroach-certs/client.root.key postgresql://root@cockroachdb-2.cockroachdb:26257?sslmode=verify-full&sslrootcert=/cockroach-certs/ca.crt&sslcert=/cockroach-certs/client.root.crt&sslkey=/cockroach-certs/client.root.key
EOF
- Run TPC-C for 30 minutes.
cockroach workload run tpcc \
--warehouses=2500 \
--ramp=1m \
--duration=30m \
$(cat addrs)
Example Output:
_elapsed_______tpmC____efc__avg(ms)__p50(ms)__p90(ms)__p95(ms)__p99(ms)_pMax(ms)
1800.0s 28185.8 87.7% 2407.9 570.4 7247.8 15032.4 28991.0 103079.2