Everserve Frequently Asked Questions

• What is Everserve?
• Can I use a different EJB?
• Will Everserve run on AIX, IRIX, or BSD?
• Can I use my own Java VM (JRE)?
• What environment variables do I need to setup with installation to run Everserve?
• Why doesn't Everserve install?
• Can I run Everserve under any account?
• Will Everserve work on the internet through a firewall?
• Can Everserve be part of a disk image when the hostname is unknown at the time the image is created?
  

How to Questions:

• How can I tell if Everserve is running?
• How do I confirm that a package was executed correctly on a remote peer?
• How can I tell which communities a peer belongs to?
• How can I exclude some files from a directory recursion?
• How do I force a remote peer to restart Everserve?
• How do I disable a remote peer that has become suspect?
• How do I create a package (specification)?
• How do I start the web based graphical user interface?
• How can I get Everserve to execute from a specific directory on the target?
• How do I rollback a package?
• How do I batch deliver Everserve packages?
• How do I schedule deliveries?

Running Everserve

• Can I reuse a package specification?
• Can a relay read the contents of packages?
• Can I start Everserve as a Service?
• Can I cancel a delivery to a peer that is permanently offline when it is pending?
• Everserve won't accept and deliver my package-specification, why?
• Everserve is not starting, what's wrong?
• Can I create sub-communities?
• Can Everserve ask the user if they want to execute the package before actually doing it?
• Can we run hidden, such that users don't even know our software is running?
• Can we run with sufficient administrative privileges to do an install?
• Can Everserve force a machine reboot?
• Can a package install the latest MS Office on Windows targets?
• Can we handle disconnected users, such as salespeople?
  

---

Answers

1. What is Everserve?

Everserve can be used to securely distribute content to mass numbers of devices, regardless of file type or size. Everserve employs a unique method of persistence, ensuring deliveries to other Everserve devices are completed even if a network or Internet connection is lost temporarily.


2. Can I use a different EJB?

Only the open source Jonas software is supported in this release. 


3. Will Everserve run on AIX, IRIX, or BSD?
   Everserve is not tested or supported to run in these environments. 
   
   
4. Can I use my own Java VM (JRE)?
   Any 1.3.1 (or better) JVM should work with Everserve. The installation program will prompt you to choose a VM or install the virtual machine bundled in with the current release. At this time, we've thoroughly tested Everserve with the JRE included on our distribution CD.
   
   
5. What environment variables do I need to run Everserve?
   Everserve does not require environment variables.  The install program does not create any new variables.  It simply modifies the path to include the Synchron/bin directory. 
   
   
6. Why doesn't Everserve install?
   Everserve 's installation program is run with a Java VM.  Java requires that the video driver support a minimum of 256 colors.  If your video driver is the default VGA driver, or the properties of the display are set to have less than 256 colors, then the installation interface will not work.
   
   
7. Can I run Everserve under any account?
   On Solaris, Everserve must run as root. On a Windows system the installation program will allow you to choose to run as another end user. It runs as "System" by default although the Services Panel allows you to reconfigure the account settings at any time.
   
   
8. Will Everserve work on the internet through a firewall?
   Yes! By configuring Everserve devices to use a dedicated JMS server located in the DMZ. The community managers and publishers are located behind the corporate firewall and communicate to externally located devices using this community JMS server. Any device that is a member of an Internet Everserve community communicate with the community managers, publishers, and relays using this community JMS server.
   
   
9. Can Everserve be part of a disk image when the hostname is unknown at the time the image is created?
   Yes, you can schedule a silent install of Everserve that executes the first time a newly imaged system is started. Alternatively, Everserve can be fully pre-installed on the disk image along with a script to patch the hostname into the appropriate configuration files once it is assigned. The Everserve community seed could also become part of the disk image with a join command set to run with the hostname auto-configuration of a newly imaged machine.
   
   
   

---

1. How can I tell if Everserve is running?
   You can type "Everserve" into a command shell.  If it changes the shell prompt and starts the interactive shell then you know Everserve is running.  If not, the command line will report that you need to run the Everserve start command. 
   
   
2. How do I confirm that a package was executed correctly on a remote peer?
   Review the receipt details on the publisher using the "everserve show receipts" command or the packages page of the Everweb application running on the publisher.
   
   
3. How can I tell which communities a peer belongs to?

List the details for the peer: 
$ everserve show communities -p <peername>


4. How can I exclude some files from a directory recursion? 

The only way to accomplish this is to create multiple directory-spec attributes in a package-spec explicitly naming the directories that are to be copied. 


5. How do I force a remote peer to restart Everserve?

Create a Package Specification that uses the AT command to stop and start all peers in the community.  You can find sample packages in the examples documentation on this portal. 


6. How do I disable a remote peer that has become suspect?

The remove peer command, executed by an administrator on the community manager peer will immediately disable future deliveries to the peer. A second command, delete peer, can permanently remove any record of the peer on the publisher and community manager systems.


7. How do I create a package (specification)?
   Use the Everweb application's packages page functionality with your browser. There is a function for editing the raw XML for adding tasks with the Everweb program. Another option is to plug the package_spec.dtd file found in Everserve's Packages directory with your favorite XML editor. Any word processor can be used to customize the package spec template files found in the Everserve/server/Packages directory.  
   
   
8. How do I start the web based graphical user interface?
   The Everweb Service must be running on the publisher or community manager to be used. You can start Everweb by choosing Start | Everserve | Everweb Session on windows devices, or by typing the following URL into a browser using the hostname of the publisher or community manager system. 
   http://<hostname>:8443/everserve
   
   
9. How can I get Everserve to execute from a specific directory on the target?
   Each spec element starts from the server directory. If you want to execute both commands on the same command try using a semicolon on UNIX or the & symbol to separate commands. A shell or batch program can also be used.
   
   
10. How do I rollback a package?
    In our current release, you implement rollback by sending another package that undoes the effects of the first package.
    
    
11. How do I batch deliver Everserve packages?
    You can implement this in three different ways: You can deliver a package to named subsets of targets within a community; you can divide targets into separate communities and send the package to one community at a time; or, you can create any number of relay peers on one physical relay server, and send the package to one named relay at a time, which will then send the package to all the targets under the relay.
    
    
12. How do I schedule deliveries?
    On Windows, you run the NT scheduling service, Task Scheduler, to set up a delivery using the AT command. You can use cron on UNIX systems. Refer to the samples and examples page of this portal for more information.
    
    

    ---

    
    

1. Can I reuse a package specification?

Yes. There is no limit to the number of times a package specification may be used. Each use will generate a unique (UUID) identifier in the Everserve system. 


2. Can a relay read the contents of packages?
   Yes. A relay can read and execute the package contents when it is configured with two peer names as both a target and a relay in the same community.  A target peer must be created with the same hostname and added to the community for this to work.  If a target  peer is not created, the relay will not execute the package; it will only pass it on to select targets. 
   
   
3. Can I start Everserve as a Service?

Yes, the install program will configure Everserve to start as a Service automatically following reboot. 


4. Can I cancel a delivery to a peer that is permanently offline - it is still pending?
   Removing the peer from the community is important so that future deliveries do not tie up resources for that peer. A utility for adding a receipt to a delivery for a peer that is permanently offline or has lost its store file is available through Synchron Technical Support.
   
   
5. Everserve won't accept and deliver my package-specification, why?
   There is probably some problem parsing the XML of the package. Everserve uses a DTD file to verify that the XML rules are followed. You may want to check the syntax by viewing the package in the Everweb package editor.
   
   
6. Everserve is not starting, what's wrong?
   First verify that the database is running if you have a problem with a community manager, relay or publisher system. Look at the logs found in the Everserve/server directory and contact Synchron's Technical Support staff.
   
   
7. Can I create sub-communities?
   This is easily accomplished by using relays. You can deliver a package to all the targets under one relay.
   
   
8. Can Everserve ask the user if they want to execute the package before actually doing it?
   You can do this if Everserve is running with permission to interact with the desktop set to yes in the Services panel. A package can put up a dialog asking the user if they want to install. If they answer no, the dialog can return a failure return code, which will end package execution. You can identify which users did not complete the install by analyzing the return receipts. Synchron Technical Support has sample Java dialog components that you can use in packages.
   
   
9. Can we run hidden, such that users don't even know our software is running?
   Yes, the startup is automatic with system boot. You decide whether you want to reveal the fact that Everserve is running.
   
   
10. Can we run with sufficient administrative privileges to do an install?
    The default account running Everserve following install is the "System" account on Windows and "root" on Unix. These accounts have administrative privileges sufficient to install a product. The account setting can be changed to any valid user of the system running the software. Refer to the System Administrator's Guide for the details.
    
    
11. Can Everserve force a machine reboot?
    Yes, Everserve ships with a program which, when invoked, reboots a Windows machine. This can be called from any package. If running as root on a Unix machine a command-spec can call shutdown or reboot.
    
    
12. Can a package install the latest MS Office on Windows targets?
    Yes, a package can send the files and commands needed to execute most any third party installation program. The Synchron Technical Support staff can provide example package specifications demonstrating how to wrap third party software installers, Operating System upgrades, etc.
    
    
13. Can Everserve handle disconnected users, such as salespeople?
    Since Everserve runs as an NT service, a portable device can connect to the specified publisher(s) or relay automatically when dialup or any type of network connection is made. The checkpoint restart feature makes this communication very efficient. If an Everserve client disconnects in the middle of a download, the download will resume from the point where disconnected when connectivity is restored.