1.4 Determining the operations policy
Determines to which configuration the execution environment is to be set to and operated at the time of building and setting up an execution environment (HCSC server).
The configuration of the execution environment differs depending on the relationship between the number of HCSC servers on which HCSC components requesting the execution of service components are deployed and multiple HCSC servers. The configuration of the execution environment and operation policy is determined on judging the scale and content of the business, the constraints of hardware that can be prepared, whether or not the migration from existing systems is easy, to what extent can the performance and reliability due to the load balancing be ensured.
In the execution environment of Service Platform, you can perform the following configurations and operations as per the relationship of the number of HCSC servers and multiple HCSC servers.
-
Operations through a single HCSC server
-
Operations through multiple HCSC servers
-
Operations configuring the load-balancing cluster in multiple HCSC servers
-
Operations configuring the HA cluster in multiple HCSC servers
-
Operations configuring the N-to-1cluster in multiple HCSC servers
In the operations through multiple HCSC servers, and operations that configure the cluster in HCSC multiple servers, you must match the SOAP mode of the HCSC server. For SOAP mode, see " 2.2.3 Used SOAP mode".
For HCSC server and cluster setup, see "3. System Setup and Unsetup".
The configurations (operations) listed above are described in the subsections of this section.
- Important note
-
If you correct the system time by using NTP software #, note the following points:
-
Execute the time correction in slew mode to gradually correct the deviation of the time to avoid time-skip.
-
If you need to do significant time correction of several tens of seconds or more at a time, stop the process of the product and then execute the time correction.
-
If you advance or return the system time due to time correction, this might cause a timeout of transaction, communication, or system monitoring transactions that becomes faster or slower than assumed. Also, incidents such as collapse of the time-series of logs and history, or duplication of the unique ID serialized in the system might occur. Review these points, and then create a system design and time-out design.
- #
-
Indicates the ntpd command, ntpdate command of Windows Time service and UNIX-based platforms.
For details on system time correction, see the manual of the OS that you are using.
-