For more information, review the topics in the Index and Table of Contents to the left. If they are hidden, click <--Show Navigation Pane.
When an Agent links to a Job for the first time, the Agent must perform an initial replication of the Job's data by downloading all of the Job's files from the Server. It will put the files in place during the linking process and start streaming their contents to the Job in the background, so the files are available to users. If someone opens a file whose information has not yet downloaded, it will focus the download bandwidth to stream that file as fast as possible, and then resume background downloads for the rest of the data.
Still, this initial replication can take a lot of bandwidth if the Job is large, and access to large files can be slow if they have to be streamed as they are opened. The FastMerge tool allows you to use an alternate copy of the Job's data to populate the linked folder initially. For instance, it may be convenient to have the original Agent write it to a DVD-R or USB drive, formatted as NTFS, and ship that to the other Agents to use for their original replication.
 |
The disk that the files are transferred with must be formatted as NTFS. If you format the disk with FAT32, FastMerge will not transfer the files properly, since the file system time stamp is local time instead of Greenwich Mean Time (GMT). You should disable thumbnail caching on Windows workstations in a WAFS/CDP environment. |
Of course, if the Job has been changed between the creation of the alternate source and the FastMerge replication, it will recognize those changes and update either the new Agent's local copy or the Job's version, depending on the rules described below. In other words, either the Agent's version will override the Job's, or the Job's will override the Agent's; either way, every computer linked to the Job will remain synchronized.
Frequently Asked Questions
|
What is FastMerge? |
FastMerge acts as an accelerator to the procedure of adding a computer to an existing Job. You begin by following the standard procedure of installing an Agent at each site and then use FastMerge to complete the initial data replication using a local copy of the data. FastMerge is effective when the content of the local folder is the same or almost the same as the content of the corresponding Job. During the time that FastMerge is running, you should make sure that the local copy of the data is not modified; otherwise, the FastMerge may result with incorrect initial replication. |
|
Why FastMerge? |
The initial mirror of data from one location might take a long time in cases of a large amount of data, especially if combined with a high latency and/or a limited bandwidth network. If the data is already available at the remote sites, or can be made available on the remote computers, you can use that data locally to perform an initial population instead of downloading it over the WAN using WAFS. FastMerge is a utility that reads the existing data from a local location and inserts it into the Agent. It assures coherency with the data that may already be in the remote Agents, and sets up the metadata for the ongoing mirroring that takes place after this initial merge is done. The standard way to move data is to install an Agent on a Windows Server at all locations and then connect each one to existing Jobs. The files appear in seconds on the other computers - immediately accessible and useable even if they are still transferring over the network. Any selected file is streamed in immediately. FastMerge replaces the transfer over the network by copying local files. |
|
When can I use FastMerge? |
You should consider using FastMerge in extreme cases of very large amounts of data, combined with long distance, which may cause the initial mirroring (via WAFS) to take a long time. FastMerge can be used for data that is already mirrored on the site, or you can transfer the data (media, FTP). FastMerge can be used in the following configurations:
If you have a large amount of data that you plan to use with WAFS, FastMerge is a utility that expedites the process of getting the data for the initial mirror into WAFS at all the remote locations. If the data is already at the remote site or you have an expedited method of transfer, WAFS will merge the data that is already local, instead of downloading all the files over the network. The data at each site does not have to be exact. FastMerge will determine and merge the differences. |
|
How long does FastMerge take? |
FastMerge execution time is the same order of magnitude as the time that it would take to do a local copy of the dataset (i.e., copy the folder to another folder on the same computer). |
|
What is the definition of "the files are the same"? |
FastMerge compares the time stamp and size. |
|
What if the time stamp is a little off, because the two computers are not in sync? |
FastMerge has a window in time in which it considers the time stamp to be the same. By default, this window is 60 seconds. The default can be changed by selecting Advanced > Managing Time Differences. |
|
Can I see ahead of time which files FastMerge is going to mirror (i.e., upload)? |
Yes. After Scan and Check is complete, you can click Show Audit File to view the FastMerge audit file, which contains a full list of what FastMerge is going to do, and why. You can share this file with users. The file is located in C:\AVCStuff\FastMerge_logs\. |
|
Can I see retroactively which files FastMerge decided to mirror and why? |
Yes, the audit file is kept, and can be viewed at C:\AVCStuff\FastMerge_logs\. There is a folder for each Job. |