Let us take a closer look at the JDBC failover connection string SQL server. At Bobcares our MSSQL support services can give you a detailed note on the process.

JDBC Connection Failover: String SQL Server

While using the default connection setup, a client application trying to connect to a down host in the Vertica Analytic Database cluster fails. This failure results in returning an error to the user.

The user must either wait for the host to recover before retrying the connection. Or manually change the connection settings to select another host.

We normally don’t care which database host handles a client application’s connection because of Vertica Analytic Database’s distributed design. When the host shown in the connection settings is unavailable.

In this case, use the client driver’s connection failover capability to avoid the user from receiving connection failures.

It provides two options for allowing the client driver to attempt to connect to a different host. Here is the one supplied in the connection parameters is unavailable. Go through all of the parameters for JDBC failover connection string SQL server.

• Set the DNS server to return numerous IP addresses for a given host name. When we use this host name in the connection settings, the client attempts to connect to the first IP address. This IP address will be returned by the DNS query.

  If we can reach the host at that IP address, the client attempts to connect to the second IP address. And so on until it either connects to a host or runs out of IP addresses.
• If the primary host given in the connection settings is unavailable, provide a list of backup hosts for the client driver to attempt.

The failover mechanism is transparent to the client application in both cases. If the primary host is not available, the client driver will attempt to connect to additional hosts automatically.

Only the initial creation of the client connection is affected by failover. If the connection fails, the driver does not attempt to rejoin to another host in the database automatically.

Choosing a Failover Method

We will have the option of using one of the two failover mechanisms. They do, however, collaborate. If the DNS server sends numerous IP addresses and we provide a list of backup hosts, the client attempts all of the IPs given by the DNS server first, followed by the hosts on the backup list.

The client does not check all IP addresses if a host name in the backup host list solves to multiple IP addresses. It just attempts the first IP address in the list.

The DNS failover solution centralizes client setup failover. By changing the DNS server settings, we can add extra nodes to the Vertica Analytic Database cluster and add them to the failover list.

All client systems that connect to Vertica Analytic Database via the DNS server automatically apply connection failover without the need for any setup changes. This solution, however, uses administrator access to the DNS server, which all of the clients uses to connect to the Vertica Analytic Database cluster. In the organization, this may not be possible.

It is simpler to use the backup server list than to modify the DNS server settings. It does, however, circulate the failover feature. If we make any changes to the Vertica Analytic Database cluster, we have to update the application settings on each client machine.

Using DNS Failover

To employ DNS failover, setup the DNS server to translate a single hostname to several IP addresses of computers in the Vertica Analytic Database cluster. We can then tell all client apps to connect to Vertica Analytic Database using this host name.

We may tell the DNS server to return as many IP addresses as we wish for the host name. In smaller clusters, we can direct it to return the IP addresses of all hosts in the cluster.

However, for bigger clusters, we should consider returning only a subset of the hosts. Otherwise, it will take for the client driver attempts but won’t connect to each server in a down database.

Using the Backup Host List

The next step in JDBC failover connection string SQL server is knowing about using the Backup host list. The client application must include at least one IP address or host name in the BackupServerNode option to allow backup list-based connection failover.

Optionally, an colon and a port number follow the host name or the IP address. If no port number is given, the driver uses the standard Vertica port number (5433). To list several hosts, use a comma to separate them.

Setting the BackupServerNode connection option to indicate additional hosts for the connection attempt is demonstrated in the following example. The connection string has a non-existent node on purpose, causing the first connection to fail. In order to connect to Vertica, the client driver must first attempt the backup hosts.

import java.sql.*;
import java.util.Properties;

When executed, the example produces the following output on the system console:

Connected! Connected to node v_vmart_node0002

Take note that the connection will be set up with the first node in the backup list.

[Need assistance with similar queries? We are here to help]

Conclusion

To sum up we have now gone through the JDBC failover connection string SQL server. With the support of our MSSQL support services, we have learned more about it.