Start YB-TServers

Note

The number of nodes in a cluster running YB-TServers must equal or exceed the replication factor in order for any table to get created successfully.
For running a single cluster across multiple data centers or 2 clusters in 2 data centers, refer to the Multi-DC Deployments section.

This section covers deployment for a single region or data center in a multi-zone/multi-rack configuration. Note that single zone configuration is a special case of multi-zone where all placement related flags are set to the same value across every node.

Example scenario

Create a 6-node cluster with replication factor of 3. - YB-TServer server should on all the six nodes, but as noted in the previous section, the YB-Master server should run on only three of these nodes. - Assume the three YB-Master private IP addresses are 172.151.17.130, 172.151.17.220 and 172.151.17.140. - Cloud will be aws, region will be us-west and the 3 AZs will be us-west-2a, us-west-2b, and us-west-2c. Two nodes will be placed in each AZ in such a way that 1 replica for each tablet (aka shard) gets placed in any 1 node for each AZ.
Multiple data drives mounted on /home/centos/disk1, /home/centos/disk2

Run YB-TServer with command line flags

Run the yb-tserver server on each of the six nodes as shown below. Note that all of the master addresses have to be provided using the --tserver_master_addrs flag. Replace the --rpc_bind_addresses value with the private IP address of the host as well as the set the placement_cloud,placement_region and placement_zone values appropriately. For single zone deployment, simply use the same value for the --placement_zone flag.

$ ./bin/yb-tserver \
  --tserver_master_addrs 172.151.17.130:7100,172.151.17.220:7100,172.151.17.140:7100 \
  --rpc_bind_addresses 172.151.17.130:9100 \
  --start_pgsql_proxy \
  --pgsql_proxy_bind_address 172.151.17.130:5433 \
  --cql_proxy_bind_address 172.151.17.130:9042 \
  --fs_data_dirs "/home/centos/disk1,/home/centos/disk2" \
  --placement_cloud aws \
  --placement_region us-west \
  --placement_zone us-west-2a \
  >& /home/centos/disk1/yb-tserver.out &

For the full list of configuration flags, see the YB-TServer reference.

Note

The number of comma-separated values in the --tserver_master_addrs) flag should match the total number of YB-Master servers (or the replication factor).

Run YB-TServer with configuration file

Alternatively, you can also create a tserver.conf file with the following flags and then run the yb-tserver with the --flagfile) flag as shown here. For each YB-TServer server, replace the RPC bind address flags with the private IP address of the host running the YB-TServer server.

--tserver_master_addrs=172.151.17.130:7100,172.151.17.220:7100,172.151.17.140:7100
--rpc_bind_addresses=172.151.17.130:9100
--start_pgsql_proxy
--pgsql_proxy_bind_address=172.151.17.130:5433
--cql_proxy_bind_address=172.151.17.130:9042
--fs_data_dirs=/home/centos/disk1,/home/centos/disk2
--placement_cloud=aws 
--placement_region=us-west 
--placement_zone=us-west-2a

$ ./bin/yb-tserver --flagfile tserver.conf >& /home/centos/disk1/yb-tserver.out &

Set replica placement policy

Note

This step is required for only multi-AZ deployments and can be skipped for a single AZ deployment.

The default replica placement policy when the cluster is first created is to treat all nodes as equal irrespective of the --placement_* configuration flags. However, for the current deployment, you want to explicitly place one replica of each tablet in each AZ. The following command sets replication factor of 3 across us-west-2a, us-west-2b, us-west-2c leading to such a placement.

On any host running the yb-master, run the following command.

$ ./bin/yb-admin \
    --master_addresses 172.151.17.130:7100,172.151.17.220:7100,172.151.17.140:7100 \
    modify_placement_info  \
    aws.us-west.us-west-2a,aws.us-west.us-west-2b,aws.us-west.us-west-2c 3

Verify by running the following.

$ curl -s http://<any-master-ip>:7000/cluster-config

And confirm that the output looks similar to what is shown below with min_num_replicas set to 1 for each AZ.

replication_info {
  live_replicas {
    num_replicas: 3
    placement_blocks {
      cloud_info {
        placement_cloud: "aws"
        placement_region: "us-west"
        placement_zone: "us-west-2a"
      }
      min_num_replicas: 1
    }
    placement_blocks {
      cloud_info {
        placement_cloud: "aws"
        placement_region: "us-west"
        placement_zone: "us-west-2b"
      }
      min_num_replicas: 1
    }
    placement_blocks {
      cloud_info {
        placement_cloud: "aws"
        placement_region: "us-west"
        placement_zone: "us-west-2b"
      }
      min_num_replicas: 1
    }
  }
}

Verify health

Make sure all YB-TServer servers are now working as expected by inspecting the INFO log. The default logs directory is always inside the first directory specified in the --fs_data_dirs flag.

You can do this as shown below.

$ cat /home/centos/disk1/yb-data/tserver/logs/yb-tserver.INFO

In each of the four YB-TServer logs, you should see log messages similar to the following.

I0912 16:27:18.296516  8168 heartbeater.cc:305] Connected to a leader master server at 172.151.17.140:7100
I0912 16:27:18.296794  8168 heartbeater.cc:368] Registering TS with master...
I0912 16:27:18.297732  8168 heartbeater.cc:374] Sending a full tablet report to master...
I0912 16:27:18.298435  8142 client-internal.cc:1112] Reinitialize master addresses from file: ../tserver.conf
I0912 16:27:18.298691  8142 client-internal.cc:1123] New master addresses: 172.151.17.130:7100,172.151.17.220:7100,172.151.17.140:7100
I0912 16:27:18.311367  8142 webserver.cc:156] Starting webserver on 0.0.0.0:12000
I0912 16:27:18.311408  8142 webserver.cc:161] Document root: /home/centos/yugabyte/www
I0912 16:27:18.311574  8142 webserver.cc:248] Webserver started. Bound to: http://0.0.0.0:12000/
I0912 16:27:18.311748  8142 rpc_server.cc:158] RPC server started. Bound to: 0.0.0.0:9042
I0912 16:27:18.311828  8142 tablet_server_main.cc:128] CQL server successfully started

In the current YB-Master leader log, you should see log messages similar to the following.

I0912 22:26:32.832296  3162 ts_manager.cc:97] Registered new tablet server { permanent_uuid: "766ec935738f4ae89e5ff3ae26c66651" instance_seqno: 1505255192814357 } with Master
I0912 22:26:39.111896  3162 ts_manager.cc:97] Registered new tablet server { permanent_uuid: "9de074ac78a0440c8fb6899e0219466f" instance_seqno: 1505255199069498 } with Master
I0912 22:26:41.055996  3162 ts_manager.cc:97] Registered new tablet server { permanent_uuid: "60042249ad9e45b5a5d90f10fc2320dc" instance_seqno: 1505255201010923 } with Master

Tip

Remember to add the command you used to start the YB-TServer to a cron job to restart it if it goes down.