SOLIDWORKS PDM Archive Replication Setup Guide

Archive replication is only available in SOLIDWORKS PDM Professional. It is not included in PDM Standard.

In SOLIDWORKS PDM, each client must communicate with the PDM Database server and an archive server.

• Information (metadata) about the files and configuration in PDM is sent to and retrieved from the database server. This is information like checkout status, version number, where the file has been replicated to, etc.
• The files themselves are copied to and retrieved from the archive server. Each time a file is checked in, that file is copied to the archive server as a new version of that file.

The file size is many times greater than the information about the file. Therefore, it takes more network resources and more time to copy the file to the archive server than for the information about that file to be sent to the database.

When clients and offices are scattered around the nation or the world, the action to copy/retrieve files to the archive can become extremely slow. For this reason, having an archive that is local to the office will speed up the copying process. These multiple archive servers are called “Replicated” servers. PDM can copy or replicate the files automatically between the servers in the evening when the network is not being used.

Archive Replication Explained

Archive replication occurs when the archive of a file is copied from one server to another server. PDM tracks which archive has each version of a file.

There are two types of replication: On-Demand and Scheduled.

On-Demand Replication

On-Demand replication happens at any time, at the request of the user. When a user selects a file in his local vault view and chooses to get a copy of that file, PDM will copy that file from his local archive server onto his computer. If the file hasn’t yet been copied to his local server, an On-Demand replication occurs. The file is immediately copied from a server that has the requested version of the file onto his local archive server. From the local archive server, the file is copied to his computer’s local vault view. In this type of replication, PDM determines which server to use as the source server. The PDM admin has no control over which server is used.

The drawback to On-Demand replication is that the user must wait for the file to be copied twice – once from a remote server to his local server and then from his local server to his vault view. The time to copy a file from the remote server to his local server can take some time if the network connection between servers is slow.

Another drawback of On-Demand replication is that it occurs during work hours and increases the network traffic during peak times. If this were the only type of replication, then there would be a big slowdown in the morning as users get the files they need.

Scheduled Replication

The goal of Scheduled replication is to copy needed files to the local server during non-work hours. With scheduled replication, the PDM Administrator decides when replication occurs. Another advantage of scheduled replication is that the administrator can decide which server to use as the source server. This lets the SOLIDWORKS PDM Administrator take advantage of how the company’s network is constructed. He can optimize the replication, so it occurs between servers with the best connection.

Another consideration when setting up scheduled replication is what files to replicate. The administrator can decide which folders in a vault need to be replicated to a particular site. So, if a group in one location only needs access to some files in the vault, then only those files need to be replicated. Replication is done by the top-level folder chosen, so if the needed files are in a subset of folders within the vault, then only those folders need to be replicated to that site. By only replicating what that site needs, you are reducing the time it takes to do the replication, and you are also reducing the storage requirements on that server.

Scheduled Replication Architecture Examples

Below are a few examples of scheduled replication architecture with some discussion of each.

Note: You may hear terms like Primary or Secondary referring to servers when talking about archive replication. These can be helpful terms to identify server infrastructure or usage, but PDM does not actually differentiate servers like this in any way. There is no “main” server. If you set up a replicated archive, you could easily remove the “original” without causing problems (so long as there were no unique files on that server). PDM sees no distinction between replicated archive servers, and any terms that would imply hierarchy are purely for ease of communication.

Hub and Spoke

In a Hub and Spoke configuration, there is one main server with all replication occurring directly to and from it. This might be a good setup if the connections between the main server and all other servers are acceptable. The drawback to this configuration is that the main server is very busy during replication times.

Ring

A Ring configuration might be a good configuration if your company is large with sites scattered through many time zones. Then you would replicate from one server to the next server at the end of the workday of the first server and before the workday began at the location of the next server.

Tree

In a Tree configuration, you take advantage of the connections between the servers. Replication occurs between the servers with the best connection to each other. While there might be a main server that has all the files, the entire vault should still be replicated to a minimum of one other server.

Prerequisites

Storage and Server Requirements

The below is true for both the main and replicated archive servers.

If only a subsection of the vault is replicated to the replicated server, then the hard drive could be smaller than on the main server. However, it still needs to be 3-5x the current size of the data it will be holding.

Note: All PDM servers MUST be on the same version of PDM

RAM: 16GB, unless you have more than 40-50 users, in which case 32GB may be prudent.

Processor: Intel or AMD processor with 4 cores. Speed is less important. Any decent server-grade CPU should be fine.

Disk Space: A MINIMUM of 500GB.

1. You want to have 3-5x the disk space available based on the current size of your data (whether it’s currently in a vault or not). This means if you currently have 250GB worth of CAD data, you’d want between 750GB and 1.2TB of storage available.
2. Be sure this disk space is scalable, or you can add further space later. PDM is guaranteed to grow over time, so being able to add more storage rather than moving to an entirely new server is preferable.
3. This storage should be in the form of locally mounted/native drives on the server. Removable storage drives or network storage, like a SAN/WAN, are not supported and can lead to performance degradation.
4. The disk should be an SSD with high i/o speeds. Hard drives can be used, but they tend to be less reliable in the long term. The archive will be written to (and over) very frequently. The ability to reorganize data and access it quickly is extremely important to ensure your users have good baseline performance. 

Network Connectivity/Latency Best Practices

1. All clients must be able to communicate with the SQL server and their archive server.
   1. Archive servers must be able to communicate with each other, as well as the database server and any clients connected to them.
   2. The default port for SQL Server communication is TCP port 1433.
   3. The default port for archive server communication is TCP port 3030.
2. If you use Windows login for the vault, both offices must have access to the domain users.
   1. This is not required if you set up a domain trust so that the local Windows domain users are correctly defined on the local domain controller. See the Microsoft TechNet library for configuration information.
3. Ideal latency is under 50ms, with 50-100ms being less than ideal, but still operable. Once you get to around the 150ms mark, we can no longer guarantee the software will operate normally. 150ms+ is considered an unstable/unsupported network latency, and could mean normal operations don’t work normally.
   1. This guideline applies to the various connections between all components that “talk” to each other – client <> server and server <> server.
4. If a client is not at an office when connecting to PDM, then they must be using a VPN (Virtual Private Network).

Steps to Set Up Replication

Main archive server

PDM has the following components to install:

1. Archive Service
2. PDM Database Service
   1. This is NOT SQL Server. This is a helper service that reads different tables in the SQL database and processes the contents of those tables.
   2. This component must be installed only once in your system. So, install either on the main archive server or on the database server. Do not install on the replicated server.
3. SQL Server
4. PDM Client
5. Web2 server (Directions for implementation of this feature can be found here.)

The main archive server must have the archive server installed on it. Often, the PDM database server is also installed on the main archive server, but it could also be installed on a separate server.

The PDM client is not necessary on the archive server for the archive server to work. However, having the PDM Client installed on the archive server is a tremendous help when troubleshooting, as it allows testing of PDM operation directly on the server. This helps identify whether issues are on the server, on the client, or in communication between the server and client.

If you are reading this article, most likely your company has already implemented PDM with a main archive server, and you are now looking at installing a replicated server. If you still need to install the main archive server, please review these instructions.

Installation on the Replicated Server

• The Archive service MUST be installed on the replicated server.
• The PDM Client is also very helpful to install for use in troubleshooting the setup and replication. Therefore, we will install both the Archive Server and the Client onto the replicated server.
• Create a folder that will contain the file archives.
  • Create it at the root of a large, local drive (SAN/NAS setups are possible, but not ideal or supported).
    • The archive grows over time, and you cannot remove replicated files once they’ve been added to the server without completely wiping and recreating the server. Ensure you have room to grow this drive or add on more drives later.  
    • If you need to move the archive location at some point, you can find instructions for doing so here.
  • Name the folder with a name that indicates it is the PDM archive.
• The PDM version must match the version on the archive server.
• Copy the installation software to the server that will become the replicated server.
• Run “Setup.exe” from the first folder. This will start the installation using the SOLIDWORKS Installation Manager or SLDIM.
  • The SOLIDWORKS Installation Manager will install or update any required prerequisites.

1. In the first dialog, select Install server components and then the checkbox Install SOLIDWORKS PDM Server components on this computer.
   
   
   
   
2. Select Change on the Summary page.
   
   
   
   
3. Choose the Server Product. This should be PDM Professional, since PDM Standard cannot use replication.
4. Now, check the box for Archive Server and Client.
5. Finally, choose Back to Summary.
   
   
   
   
6. Back on the Summary screen, accept the terms from SOLIDWORKS and then select Install Now.
   
   
   
   
7. PDM will perform the installation, including any required prerequisites.
8. Once completed, this screen will display. Make your choice on the SOLIDWORKS Customer Experience Improvement Program, then select Finish to exit the dialog.
   
   

Archive Server Configuration

Archive Server Configuration will run if this is the first time PDM Archive has been installed on this server. If the PDM archive server has been installed before, but an uninstall was done, and you are now reinstalling, then this next section will not run. The information is stored in the registry.

1. Once the main installation is complete, the Archive Server Configuration will begin.
2. Select Next.
   
   
   
   
3. Specify the location for the PDM Archives on this server. You may have created this folder before starting the installation.
4. If you still need to create the folder:
   1. Create the archive folder at the root of a large locally mounted drive.
5. Name the folder with a name that indicates it is the PDM archive.
6. Select Next.
   
   
   
   
7. Enter the Default Admin user password.
8. Select Next.
   
   
   
   
9. Enter the SQL credentials that PDM will use to connect to the PDM database.
10. Select Next.
    
    
    
    
11. There are two sections in the Security page: Administrative access and Attach access. All of the accounts/groups listed here are Windows users. PDM scans the Windows account logged into the PC (not PDM – they are completely separate) when performing actions. This allows you to designate which Windows users should have these permissions.
    1. Administrative access: Anyone in the administrative access section can create and destroy vaults. Be very careful who is added to this section.
    2. Attach access: Users with attach access can add vault views onto clients.
    3. If needed, these settings can be modified in the Archive server Configuration Tool after installation is completed.
12. Select Next to continue.
    
    
    
    
13. Login Type: Choose the default login type. This is likely going to be the same type you use on your existing Archive server(s).
14. Select Next.
    
    
    
    
15. Select Finish to close the Completed screen.
    
    

Attaching the Replicated Vault

With the software being installed on the replicated server, we can set up the replicated vault.

1. To do this, open the PDM Administration tool.
2. Right-click the server’s name and choose Add replicated vault…
   
   
   
   
3. Choose Next on the Welcome screen.
   
   
   
   
4. Enter the name of your main archive server. Use the default port of 3030 and choose Next.
   
   
   
   
5. Accept the license agreement.
   
   
   
   
6. This next screen can cause a lot of confusion. In this case, it showed up when we were adding a vault to this server. But it also frequently shows up when creating a vault view on a client computer.
   1. This dialog is looking for a Windows account that has been added to the Administrative access section of the Archive Server Configuration tool’s Security The Domain can be typed in.
   2. We cover Attach/Admin access more here if you need some clarification.
      
      
      
      
7. Select the vault you wish to replicate and select Next.
   
   
   
   
8. Enter the PDM admin account password for this vault and then select Log in. This will allow you to log in to the PDM admin tool.
   
   
   
   
9. This next selection can also be confusing, but PDM is asking where to place this vault in the folder structure in the Archive Server Configuration tool. It is best to use the default Archives folder.
   1. Note: This is NOT the location of the archive files on your server drive.
10. Select Finish.
    
    
    
    
11. The last dialog is just a reminder on how to start scheduled replication. Select OK to close the dialog.
    
    
    
    
12. The vault now shows in the PDM Administration tool on the replicated server.
    
    

PDM Admin Account Setup in a Replicated Environment

During the installation of the PDM Archive server, the default SQL credentials were entered, as well as the default PDM admin password. In a replicated environment, it is desired that the PDM admin password be the same regardless of which server the vault resides. To achieve this, the PDM admin password must be set to NOT use the default password. It must use the vault-specific password. This way, any change to the PDM admin password will be replicated to the other servers.

Note: This must be set on all archive servers.

1. Go into the Archive Server Configuration tool. You can search for it using the Windows Start menu, or under Start > All Programs > SOLIDWORKS PDM.
   1. If you try opening the tool and nothing seems to happen, try opening up the hidden icons from your System Tray (down near the time and date on the Windows taskbar) and clicking the PDM Archive Icon to open it instead. If it was previously opened, the tool stays running in the background when closing it unless you explicitly use the system tray to Exit.
      
      
      
      
2. Expand Archives.
   
   
   
   
3. Right-click your vault and choose Properties.
   
   
   
   
4. Select the Login tab.
5. Clear the box for Use default Admin user.
6. Verify the box is checked for Use default SQL login.
   
   
   
   
7. When done, click OK to apply changes and close the Archive Configuration tool.

Setting Up Scheduled Replication

Now we’ve got the two servers connected and the vault ready for replication between both. If you have a PDM client connected to your new replicated server and they try to access any files in PDM, they will replicate on-demand. Many companies like to set up scheduled replication to move files automatically during off-hours so that users don’t have to wait for them to copy over when they go to access the file. That way, the file is already cached on the local server when they go to access it. To do this, we’ll need to set up a replication schema. With this, we can choose what folders and versions of files to replicate, to/from which servers, as well as the time we want them to replicate.

Open the PDM administration tool. Log into the vault that is now being replicated. Expand Replication > Archive and double click Replication Settings.

This takes us to the Replication Settings dialog.

Here’s a quick rundown of the various parts:

1. Test column is a communications test from the Source Server to the Destination Server. You can check the boxes for the servers and direction you want to test. The Test button brings up the test dialog box once you select which connections you’d like to check.
2. Target Server is the server the files are being copied to. Scheduled replication is a Pull. This means it is the Target Server that triggers the replication.
3. Source Server is the server the files are being copied from. One advantage of scheduled replication over On Demand replication is that the admin can specify exactly which server to do the replication from.
4. Source Server Address: Enter a valid hostname or IP address that the Target Server can use to communicate with Source Server.
   1. In a WAN environment, you should use the IP address here. The archive servers will communicate over TCP port 3030, so make sure that port is fully opened in any firewall between the servers.
   2. Select the row, and then the Source Server Address and Priority can be edited in (6).
5. Priority: Balance the replication load between three or more servers. Server connections with a higher speed rating take priority over servers with a lower speed rating. If a file version is replicated on two servers and a third server requests that file version for replication, the connection that has the highest priority setting is used first since it exists on both remote servers.
   1. Priority is really only important when you have 3+ replicated servers. If you are replicating between two servers, you should select the highest speed rating (10 Mbit) regardless of real connection speed. The speed setting does not affect the actual replication transfer speed in any way, only priority.
6. Selected connection is the location to edit the Source Server Address and Priority level from (4) and (5).
7. Replication Schemas: All scheduled replication is set up and managed with a replication schema. This area shows the existing replication schemas. They can be edited by selecting one and choosing Edit Schema.

Test the Connection Among the Replicated Servers

The first step in setting up replication is to verify the servers can talk to each other. PDM can do a communications test.

In the Replication settings dialog box, select the checkboxes under the Test column, then select Test.

When the Test Connections dialog box appears, notice their Status is Waiting. Select Test to start the test.

Once the test has been completed successfully, note the Status column has changed to Succeeded, and the Result column says, Test successful. If it fails, the Status column will say Failed, and the Result column will provide some clarity as to why.

If the test fails, the most common reason is that port 3030 is blocked between the servers, or there’s an issue with name resolution or TCP/IP communication between them. Be sure to check that the name/IP specified in the selected connection section mentioned earlier is pointing to the correct location.

Set Up a Scheduled Replication Schema

Now that the connections between servers have been successfully tested, it is time to set up the Scheduled Replication.

1. Select New Schema in the Replication settings dialog box.
   
   
   
   
2. When scheduled replication starts, PDM will replicate the Folders that are part of the schema from the Source Server to the Target Server.
   1. Look closely at the column headings and note the Target Server is on the LEFT.
3. Give the schema a descriptive name, so that you can tell what is being replicated and in what direction just from the name.
4. Next, select Add Folder… to add folders to be replicated. You can choose to schedule replication for a select number of folders, or the root of the vault to replicate all files.
   1. Note that folders selected here only designate which folders/files are replicated on a This does not designate permissions/access. If a user has proper folder and state permissions, they will still be able to access and replicate files on-demand outside of the folders you choose here.
      
      
      
      
5. Selecting Add Folder brings up the Select Folder dialog. This allows you to browse through your folders to select the folders you wish to replicate in this schema. In this case, I am selecting the entire vault.
   
   
   
   
6. The selected folders are listed in the Folders that are part of this schema section.
   1. Note: If you make a mistake, you can select the folder and choose Remove Folder to remove it from this schema.
      
      
      
      
7. Now we can set up the replication schedule. Select a connection from the list for replication. Pay close attention to which is the Source Server and which is the Target Server.
   1. When there is no Interval defined, then no scheduled replication will occur, only on-demand.
   2. The times listed will be the Target Server local time. So, if your source server is Eastern time, and your target server is Pacific, 00:00 would be replicating at midnight Pacific time and 3:00 AM Eastern time.
8. Choose a predefined time from the list or select User defined schedule to set custom replication times.
   1. If you already have a scheduled interval and want to remove it, then select the interval to remove and select No schedule.
      
      
      
      
9. In the example below, files will be copied from Jward-2 to JimWard-Laptop at 6:00 PM (6:00 PM for JimWard-Laptop’s time zone) every day of the week.
   1. Note the format is minute hour weekday.
      • Minute starts at 0 and goes to 59. “0” means the top of the hour.
      • Hour starts at 0 and goes to 23.
      • The days are numbered 1-7, with 1 being Monday.
      • An "*" means every interval. So, an "*" for minute would mean to run the schema every minute of the specified hour.
10. Usually, replicating once per day is enough. Remember that On-demand replication always works, so if a user needs a file that has not yet been replicated, then on-demand replication will take care of the replication.
    1. Never schedule replication more than 2-3 times per day maximum. There is no benefit to doing this more often, but there are definitely drawbacks. Scheduling replication too frequently can cause performance impacts and instability for your end users.
    2. If you are replicating the same folders in both directions, you can set up the replication schema with a replication interval in both directions. Then, you don’t need a second schema.
       
       
       
       
11. Once the interval is set, the replication will begin at the specified times. There are some further options for the replication that are found under the Advanced selection at the bottom left of the dialog.
    
    
    
    
    
    
12. Let’s review the options:
    1. Replicate only latest version: Replicates only the most recent version if a file has multiple versions that are not yet replicated to a server.
       • For example, if versions 5 through 10 are not yet replicated to server B, enabling this option replicates only version 10. Without this option enabled, versions 5 through 10 are replicated when the schedule runs.
    2. Replicate only files with revisions: Replicates the file versions that have a SOLIDWORKS PDM revision assigned.
       • For example, if version 4 has revision A and version 7 has revision B, enabling this option only replicates versions 4 and 7 when the schedule runs.
    3. Include children when replicating: By default, only files stored under the specified folders set up in the schema are replicated by the schedule. Any files not under the specified folders are not replicated automatically. With this option enabled, the schedule also includes all referenced files to a file stored under the specified folder of the schema.
       • For example, consider a schema that is set to replicate only files under the \Projects folder. The vault contains an assembly file that references two part files under the \Library folder. With this option enabled, those two part files will also be replicated when the schema is run, even though the \Library folder is not part of the replication schema.
    4. Only replicate files in the following states: Includes only files that are currently in the specified workflow states. Select a workflow and state in the drop-down list and click Add. Any file versions not in the specified state are not replicated.
    5. Stop replication after: By default, once a scheduled replication starts, it processes all files and versions not yet replicated to the server. The files are added to a queue and are replicated until the queue is empty. To stop the replication process after a set length of time, you can enable this option and specify when the replication should stop. Any files not yet processed after that period are added to the queue for replication the next time a schema is set to start.
       • If a client retrieves a modified file before the next time the schema runs, the file is immediately replicated by on-demand replication.
       • This option can be useful when first setting up a new replicated server. Without this, the replication will continue until all files have been replicated, which might be days if there are 100s of thousands of files to replicate.
13. When finished, select OK to go back to the previous screen and save your changes.
14. Select OK on Edit Replication Schema to return to the previous screen.
    
    
    
    
15. In the top left of the PDM Admin tool, select the Save button to save your Replication settings changes. You can also click the red X in the upper right for a prompt to save, which you can click Yes to.
    
    

Archive Status

Archive Status gives you information about your replicated vault. You can see which servers are being replicated to, what percentage of files are replicated, specific file counts, and how many files should be replicated but have not yet done so.

Double-click Archive Status, and this dialog will appear. Choose the replicated server of interest at the top, and the replication status of that server will be shown.

If you’d like to see a list of files that have not yet been replicated, but should be according to your schema, right-clicking Archive Status > Show Missing Files will bring up an Excel spreadsheet that will show the files that have not been replicated to this server.

Be aware that Unreplicated files to server includes files that have been deleted but not destroyed. This is why Deleted files is listed. If “Unreplicated files” and “Deleted files" are about the same, then there is no issue.

Final Thoughts

Archive server replication is needed for companies with multiple sites spread around the country or world. With archive server replication, it is feasible to have designers contribute to your design without them being physically close to your main office.

While individual users can use Virtual Private Networks to access your main office, this is not feasible when you have another office. In this case, a replicated archive server may well help your designers to become much more productive.

If GoEngineer is your reseller and you have any questions or issues, please reach out to our Support Team for more help! We’d be happy to assist.

Editor's Note: This article was originally published in October 2016 and has been updated for accuracy and comprehensiveness.

Learn More About SOLIDWORKS PDM

SOLIDWORKS PDM System Resources Starting Point

3 Fixes for SOLIDWORKS PDM Client Crashing

How to Color Code SOLIDWORKS PDM Warnings

Deleting a Local Vault View in SOLIDWORKS PDM

SOLIDWORKS PDM Settings Dialog Recommendations

View All PDM Articles