Disconnected Server Installation and Update Guide

This a guide to installing a Disconnected Triaster server.

Overview

A Disconnected Server (DS) is a machine with a working version of Triaster Server installed, but that is not intended to publish or serve content over a network. Rather, DSs are used where for example a copy of a normal set of Triaster Process Libraries are required off-shore or in other areas where it is not possible to connect to the actual server. Users access the content through the physical machine itself, not over a network.

Triaster have built-in replication routines so it is simple to periodically update a DS with content from a live server.

Because a DS is not intended to be a server, it can be installed to a machine that is running Windows 10 rather than a Server OS.

There are 3 main steps:

  1. Configure Replication on the Live Server
  2. Install the Triaster Server software onto the DS (only required to be done once)
  3. Import the Replication content
  4. Refresh the search database

These steps are described below.

Configure Replication on the Live Server

The Replication parameters are found in the Replication Node in Settings.xml. Only skilled users should work with Settings.xml.

  • Frequency, one of Hourly, Daily or Weekly; this is how frequently the Live Server replicates
  • Timestamp, this is the date and time that controls when during replication actually happens. It is in ticks (see http://tickstodatetime.com/). For example, 636687416881847644 is August 1st 2018, 17:34. So, if Frequency is Daily, this Timestamp means Replication will occur every day as soon as possible after 17:34. As the given date is on a Wednesday, if Frequency is weekly, it will happen every Wednesday and so on. The default value is 636218560000000000 which is February 5th 2017 (a Sunday) at 01:46.
  • TargetPath, the root folder of the export
  • DeleteTargetPath, optionally set this to True to force deletion of the target path before the replication takes place. Set to False (the default) to leave the existing Replication content in place. The effect of True is to propagate deletes from the Live Server which is generally preferred. The effect of False is to retain old content but also to not cause all the content to be copied every time the replication is performed, and this is gneerally preferred when dealing with large data sets.
  • PerformReplication, set to True to turn on Replication, False to turn off-shore
  • LastReplicated, a Timestamp representing the last successful replication. Delete this to trigger a replication ASAP.
  • TargetZipFile, optionally, the contents of TargetPath can be zipped into the file specified in this field. This will cause an existing zip file to be deleted.
  • CopyOperation, optional and there can be many CopyOperation tags each containing the following tags.
    • Source
    • Target
    On each run of the Replication, the server will copy the contents of Source to Target. If Target exists it is not deleted prior to the copy.

Installation Process

  1. Obtain a machine - the ("DS") - that meets the Minimum system requirements with the following exceptions:
    • Visio is not required
    • Process Navigator is not required
    • Windows 10 or later will suffice if the OS is not Windows Server
  2. Download and unzip the latest version of Triaster Server onto the DS
  3. From the unzipped files, run Setup.exe
  4. Check the tab Minimum Requirements Compliance Report. Make adjustments if necessary (bearing in mind the above exclusions)
  5. Enter your network User Name and password on the Edit Details tab
  6. Click Start Installation
  7. Run \Triaster\TriasterServer2011\Services\PubServerConfig.exe. The steps below are the configuration settings which should be changed by the implementer.
    • All Servers tab: Change path if applicable (e.g. from C:\ to D:\)
    • Publication Server tab: change from default (Server1 to actual computer name or 'localhost' (without the apostrophes))
    • Token Server tab: change from default (Server1 to actual computer name or 'localhost' (without the apostrophes))

Import or Update DS Content

This is the process whereby a replication package that has been exported from the live server is imported to the DS. The same process is performed on the first run through and each subsequent run through.

A replication package contains the following directories found in \Triaster\TriasterServer2011 and all their contained files and sub-directories:

  • \BrowserToolkit\css
  • \BrowserToolkit\themes
  • \KeyotiSearch
  • \MenuPage Templates
  • \ProcessLibraries
  • \Reports
  • \Settings

Important! Performing this process will cause the corresponding local content in each of the above folders to be deleted.

  1. Run \Triaster\TriasterServer2011\Services\PubServerConfig.exe
  2. Select the Disconnected Server tab: change Location of Data Copy to Import From to point to the live server's export folder. Note that this might also be to a pen drive, DVD etc that has been produced by copying the replication package.
  3. If Local Machine Installation Location is incorrect, or you would prefer to test the Import by writing to a different location, change it to a different folder
  4. Click Import

The import process may take significant time depending on the amount of content to be imported. Please do not interrupt the process.

The import process first deletes the local content, and then copies the files found in the Import location into the local TriasterServer2011 folder. It then replaces any reference to the live server to localhost in each of the following files:

  1. Settings\Settings.xml
  2. ProcessLibraries\Libraries.xml
  3. KeyotiSearch\IndexDirectory\indexableSourceRecord.xml

Re-index Search

Run KeyotiSearch\KeyotiReindex.exe - this will cause the search indexes to be re-built with all the new local content.

Last Reviewed: 23/1/2019