Frequently asked questions about HFSQL Cluster
Available only with this kind of connection
A specific setup package is required to use HFSQL Cluster.
This pack can be downloaded at no extra cost on PC SOFT site, in the download section of modules common to version 28.
|Question||What is HFSQL Cluster?|
|Answer||HFSQL Cluster is a system for managing relational databases intended to be deployed on a group of servers (a cluster) in order to guarantee a high availability. For more details, see Overview of HFSQL Cluster.|
|Question||What is the link between HFSQL Cluster , HFSQL Client/Server, HFSQL Classic and HFSQL Mobile?|
|Answer||These terms are used to identify the different editions of the HFSQL database. All these editions share the same file format which makes evolutions and data transfers very easy. All the editions also share the same statements for accessing the data.|
- HFSQL Cluster: cluster edition intended for critical architecture requiring high accessibility.
- HFSQL Client/Server: Client/Server edition. The databases are made available to the client applications by a database server.
- HFSQL Classic: Local or Network edition. The databases are stored in simple files that can be used by a single application on a local computer or that can be shared on a network.
|Question||What is the license proposed for using HFSQL Cluster?|
|Answer||The HFSQL Client/Server license also applies to HFSQL Cluster.|
|Question||Does HFSQL Cluster require a specific hardware configuration?|
|Answer||HFSQL Cluster does not require any special investment to be implemented. For more details, see Installing HFSQL Cluster.|
|Question||Is HFSQL CLuster intended for a specific type of applications?|
|Answer||All types of applications can be addressed by HFSQL Cluster, especially in the fields that require high data availability and/or the fields that require load balancing (Web hosting, ...).|
|Question||Do I have to modify my existing HFSQL Client/Server code to use HFSQL Cluster?|
|Answer||No. The code is entirely compatible between the different editions of HFSQL. However, in its current version, HFSQL Cluster has a few limitations compared to HFSQL Client/Server (these limitations are listed below). It is very important to run the test of any application intended to be deployed on a cluster before it is implemented.|
|Question||How many physical server are required to create a minimum cluster configuration?|
|Answer||At least three physical servers are required to create a configuration of HFSQL cluster: two of them will be the nodes of the cluster, the third one will run the service of the Coordinator (ClusterManager service).|
|Question||What is the maximum number of nodes in an HFSQL cluster?|
|Answer||A cluster can manage up to 128 nodes.|
|Question||Can the physical servers of a cluster be located in several geographical locations?|
|Answer||Because of the throughput required to perform a synchronous replication of the databases, we highly recommend not to dissociate the nodes of a cluster geographically. However, a mechanism for asynchronous replication can be used between two clusters that are geographically distant.|
|Question||Can client applications be in 32 and 64-bit?|
|Answer||Client applications can run on Windows 32 or 64-bit, Linux (32-bit only), and MacOS X (32-bit only), regardless of the version of the cluster (32 or 64-bit).|
|Question||How to change the major version of the HFSQL engine in the nodes of an HFSQL cluster?|
|Answer||The change of the major version of the HFSQL Client/Server engine is done successively on each node:|
- Stopping a node,
- Update of the HFSQL Client/Server engine for the stopped node,
- Wait for the node to synchronize before performing the same operation on another node.
|Question||How is the status of an HFSQL Cluster monitored?|
|Answer||The status of the nodes of an HFSQL Cluster can be monitored via the Monitoring Robot.|
|Question||After changing the file structure in the analysis (adding, changing or deleting items), how to apply the changes to the files on the different cluster nodes?|
|Answer||Simply apply the structure modification (automatic data modification) to a cluster node.|
The HFSQL Cluster synchronization process transfers structure modifications to all the cluster nodes.
|Question||Is it possible to add or remove links? |
|Answer||Starting with version 27, automatic data modification support is more comprehensive: |
|Question||Does HFSQL Cluster have any limitations?|
|Answer||The current version of HFSQL Cluster has the following limitations compared to HFSQL Client/Server.|
These limitations will disappear in the forthcoming versions.
- No bounds of automatic identifier can be defined for the replication.
- A backup cannot be restored without stopping the cluster.
- The cluster is not synchronized if stored procedures or queries are found.
- The following functions are not allowed:
- HCopyFile (only the Client to Server syntax and the Server to Server syntax)
- HSetServer (the following options are not available: hLogLevel, hMaxLogSize, hActivityStatisticsPeriod, hMaxActivityStatisticsSize, hServerPort, hMode2GB and hDebuggingPort)
This page is also available for…
Click [Add] to post a comment