![[Contents]](FIGURE/CONTENT.GIF)
Job Management Partner 1/Software Distribution Description and Planning Guide
![[Glossary]](FIGURE/GLOSS.GIF)
![[Index]](FIGURE/INDEX.GIF)
![[Back]](FIGURE/FRONT.GIF)
![[Next]](FIGURE/AFTER.GIF)
When you deploy JP1/Software Distribution, the software to be installed and various types of management information are transmitted via the network. To ensure efficient operation, you should minimize the load on the network by running jobs during low-usage periods. JP1/Software Distribution provides several facilities for reducing the network load. These include adjusting network traffic during setup and distributing software to be installed in segments. This section explains how to minimize the network load and distribute software efficiently. In order to distribute software efficiently in JP1/Software Distribution, the settings and operations on the client machines must also be considered. For user tasks at the client, see 11. Using Client Facilities in the manual Administrator's Guide Volume 1.
To transfer packages efficiently during remote installation, you must understand the role of relay systems.
A relay system not only reduces the number of clients connected directly to the managing server but also efficiently transfers a package to the clients under the relay system by creating and saving a copy of the package. Also, by combining a Transfer package to relay system job with an Install package job, a package can be transferred in separate stages, first to the relay system and then beyond the relay system.
The following figure shows the operation of a relay system during remote installation.
Figure 6-5 Operation of a relay system during remote installation
![[Figure]](FIGURE/KAN0210.GIF)
A Transfer package to relay system job is stored in the relay system. While the job is stored in the relay system, the managing server does not redistribute the same package. During packaging, you can specify the storage period.
If the number of leased lines is limited, or if the network rejects access, it may be advisable to adjust the number of hosts that can be connected concurrently. The managing server provides the following methods to adjust the number of hosts that can be connected concurrently:
The managing server can use setup information to control the network traffic. From the menu, choose Setup to open the setup window, and then display the Server Customization page (for a relay system, use the Relay System Customization page):
Figure 6-6 Server Customization page
![[Figure]](FIGURE/KAN0220.GIF)
The option for adjusting the number of hosts that can be connected concurrently is Max. number of subsystems in which jobs can execute concurrently (for a relay system, it is Max. number of relays or clients in which jobs can execute concurrently).
If a managing server executes jobs for a large number of clients at the same time, the load on the network may become extremely heavy. Therefore, specifying Max. number of subsystems in which jobs can execute concurrently can limit the number of lower hosts that can be processed concurrently. For example, if a maximum of 20 is set, a job that remote-installs software in 300 clients concurrently is executed in 15 batches of 20 clients each.
During setup, you can specify values that take into account PC performance, network configuration, and operating environment. This provides sufficient reliability in most cases. However, when large amounts of data are transmitted, the network load might increase even if you specify settings to control the traffic during setup. Therefore, it may still be necessary for users to make adjustments.
Suppose that a 5-MB package is to be transmitted over a 64-Kbps line. If the line efficiency is 60%, transmission will take some 18 minutes. The package will flow in parallel between the managing server and the relay systems and between the relay systems and the clients. Determine the number of target clients for a single job by considering the load in the connection lines, the speed of the connection lines, and the line efficiency.
To avoid concurrently connecting a large number of relay systems to the managing server, we recommend split job execution. Although this method places some burden on the user, it can reliably reduce the network load.
First, by considering the network traffic, determine the number of relay systems to be connected concurrently to the managing server. Next, for each relay system that will be concurrently connected, classify all clients under each relay system as a single group. When you execute the job, leave an interval between job execution times for the different groups. In this way, you can avoid having a large number of relay systems connected concurrently to the managing server.
When you create a remote installation job, you can specify the job execution time. This is the package transmission time. By specifying a late-night hour, when traffic on the network is relatively light, data can be transmitted more efficiently.
If you use separate jobs to execute package transmission from the managing server to the relay systems and from the relay systems to the clients, you can specify different execution times for these jobs.
You can reduce network load by carefully scheduling transmissions rather than reducing the volume of data to be transmitted.
Note that software distribution is not executed if a destination PC is in the power-save mode, because its JP1/Software Distribution Client (client) is not active in that mode. If destination PCs have the power-save mode, exercise care when scheduling jobs.
Software to be installed in a client can be distributed in a single batch or in multiple segments. The latter is called split distribution. Split distribution offers the following benefits:
The following figure shows software distribution using split package distribution.
Figure 6-7 Software distribution using split package distribution
![[Figure]](FIGURE/KAN0230.GIF)
In normal job distribution, the number of packets sent from the higher system increases as the number of clients increases. JP1/Software Distribution supports a distribution method in which the higher system sends the packets for a job only once for distribution to the specified number of clients, reducing the transfer load. This is called multicast distribution.
When you specify multicast distribution for distributing software, you reduce the number of packet sends, thus reducing the time required for distribution and reducing the load on the network. Multicast distribution is effective in the following cases:
The figure below shows software distribution using multicast distribution. You can use multicast distribution only between clients and the higher system to which they are connected.
Figure 6-8 Software distribution by multicast distribution
![[Figure]](FIGURE/KAN0240.GIF)
For details about the system configuration and setup required for multicast distribution, see 6.2 Settings for multicast distribution.
You can suspend job execution temporarily at a host where package transmission or remote installation is underway. For example, if a job is to be executed while a specified application is not running, but the job is not completed when the application must start, you can suspend the job and restart it after the application has terminated.
You can also distribute one job while suspending distribution of another. For example, if it is important to distribute a virus definition file in a hurry, you can suspend the current distribution job and distribute the virus definition file first.
There are three methods for suspending and resuming a job:
For details about the dcmsusp command, see 4.23 dcmsusp.exe (suspending and resuming a file transfer) in the manual Administrator's Guide Volume 2.
The unit of file transmission that can be suspended is the value of File transfer buffer size that was specified during setup of the lower system.
Sending a suspension or resumption instruction from the managing server to the relay manager/system
To instruct suspension or resumption, execute a Suspend file transfer job or Resume file transfer job from the managing server. For the destination of such a job, specify the higher system directly connected to the hosts where the job is to be suspended or resumed.
The following figure shows the system operation during execution of a Suspend file transfer job.
Figure 6-9 System operation during execution of a Suspend file transfer job
![[Figure]](FIGURE/KAN0250.GIF)
The following figure shows the system operation during execution of a Resume file transfer file.
Figure 6-10 System operation during execution of a Resume file transfer job
![[Figure]](FIGURE/KAN0260.GIF)
Issuing a suspension or resumption instruction to the managing server itself
You can use the Remote Installation Manager to issue a suspension or resumption instruction to the local system. To do this, from the Execute menu, choose Suspend/Resume File Transfer, and then Suspend or Resume.
The following figure shows the system operation during job suspension or resumption at the local system.
Figure 6-11 System operation during job suspension or resumption at the local system
![[Figure]](FIGURE/KAN0270.GIF)
You can distribute a job without suspending file transfer even when the system immediately above the destination system is in suspended status. To do this, during job creation, open the Create Job dialog box, choose the Job Distribution Attributes tab, and then Distribute.
The following figure shows distribution of a job to a suspended destination.
Figure 6-12 Distributing a job to a suspended destination
![[Figure]](FIGURE/KAN0280.GIF)
When a job is suspended, file transfer to destination systems will be suspended while the system immediately above them is in suspended status. However, if you select the Distribute option for a job, file transfer will not be suspended.
The following shows the scope of job suspension and resumption:
JP1/Software Distribution provides a facility that enables remote PCs connected via a network to be started and shut down from the local PC. This facility is called client control. JP1/Software Distribution can use this facility to turn on remote PCs and install software in them at night or on a weekend.
To use this facility, the PC must support either AMT or Wake on LAN, and must also support automatic shutdown.
The following figure shows an overview of software remote installation using client control.
Figure 6-13 Remote installation using client control
![[Figure]](FIGURE/KAN0290.GIF)
For notes on remote installation using client control, see 6.3 Settings for using the client control facility.
You do not need to use a network to install software on a stand-alone PC (offline machine) on which JP1/Software Distribution Client is installed; this is called offline installation.
Offline installation is useful in the following cases:
Offline installation involves storing the files needed for installation on storage media at the managing server and then transporting that media to the offline machine for installation. To use offline installation at the managing server and an offline machine, the following conditions must be met:
For details about the offline installation method, see 7.7.1 Offline installation in the manual Administrator's Guide Volume 1.
All Rights Reserved. Copyright (C) 2009, 2013, Hitachi, Ltd.
Copyright, patent, trademark, and other intellectual property rights related to the "TMEng.dll" file are owned exclusively by Trend Micro Incorporated.