Use Telnet to verify whether your ApsaraDB for HBase cluster is reachable. Because ApsaraDB for HBase does not support the Internet Control Message Protocol (ICMP), ping does not work against a cluster. Use Telnet on port 2181 (the default ZooKeeper port) instead. ApsaraDB for HBase supports both intranet and Internet connections. To connect over the Internet, first apply for a public endpoint for your cluster.
Prerequisites
Before you begin, ensure that you have:
-
An ApsaraDB for HBase cluster
-
Telnet installed on your client machine
-
Network access to the cluster (intranet or public endpoint)
Get the ZooKeeper endpoint
-
Log on to the ApsaraDB for HBase console. On the Clusters page, click the ID of your cluster.
-
In the left-side navigation pane, click Database Connection.
-
In the Connection Information section, copy the endpoint for your cluster edition:
-
Standard Edition: copy the value of ZooKeeper Address.
-
Performance-enhanced Edition: copy the value of Java API Address.
-
If you connect using Thrift or another database service, use the corresponding domain name and port for that service instead.
Test connectivity with Telnet
Run the following command, replacing <zookeeper-address> with the endpoint you copied:
telnet <zookeeper-address> 2181
Connected — the cluster is reachable:
[xx@yy-MacBook-Pro ~]$ telnet hb-xxxxx-001.hbase.rds.aliyuncs.com 2181
Trying 10.10.10.10...
Connected to hb-xxxxx-001.hbase.rds.aliyuncs.com.
Escape character is '^]'.
Not connected — a timeout indicates a network or configuration issue:
[xx@yy-MacBook-Pro ~]$ telnet hb-xxxxx-001.hbase.rds.aliyuncs.com 2181
Trying 10.10.10.10...
telnet: connect to address 10.10.10.10: Operation timed out
telnet: Unable to connect to remote host
If you see Operation timed out, work through the troubleshooting steps below.
Troubleshoot connectivity failures
Whitelist not configured
Symptom: Operation timed out
Cause: The client IP address is not on the cluster allowlist.
Fix: Configure a whitelist to allow access from your client. See Configure a whitelist and a security group.
ApsaraDB for HBase cannot reach the client
Symptom: Operation timed out when testing from an Elastic Compute Service (ECS) instance.
Cause: The ECS security group or firewall blocks inbound connections from ApsaraDB for HBase.
Fix: Update the ECS security group rules to allow inbound traffic from ApsaraDB for HBase.
Client is on Classic Network, cluster is in a VPC
Symptom: Connection fails despite correct endpoint and whitelist.
Cause: Classic Network clients cannot connect to ApsaraDB for HBase clusters deployed in a virtual private cloud (VPC).
Client and cluster are in the same VPC but on different vSwitches
Symptom: Operation timed out between resources in the same VPC.
Cause: Different vSwitches do not route traffic to each other by default.
Fix: Configure routing rules between the vSwitches.
Client and cluster are in different VPCs
Symptom: Operation timed out across VPC boundaries.
Cause: Two different VPCs are isolated from each other by default.
Fix: Use Express Connect to establish communication between the two VPCs.
Client is outside Alibaba Cloud (for example, your company's private network)
Symptom: Operation timed out or Unable to connect from an on-premises host.
Cause: The client has no network path to the cluster's intranet address.
Fix: Choose one of the following:
-
Apply for a public endpoint for the cluster and connect over the Internet.
-
Use a leased line to connect your on-premises network to an Alibaba Cloud data center.
ECS instance cannot connect over the Internet
Symptom: Internet connection from an ECS instance fails.
Cause: The ECS instance does not have Internet access enabled.
Fix: Enable Internet access on the ECS instance, then retry the connection.
If the connection still fails after working through the steps above, contact the ApsaraDB for HBase expert team on DingTalk (DingTalk ID: s0s3eg3).