I saw the project Hive2Hive, I think it is a very good project and I am very interest. I am involved with a project, that has to save the files of different applications in a File System just like this one. See image here
Let me describe the structure of this project: There are 2 Applications in 2 different servers.
The Agent:
is responsible to send the content of a folder to a Collector This content is a snippet of the file
only sends information if the Collector asks
- it listens the Collector and sends the content by the same door
after sending it, might delete or truncate the file or o nothing else
it needs the key(.ssh/id_rsa) and if necessary a user With the key, it can do an context switch or user switch, thus being able to delete or truncate the file
- it communicates with only a Collector
it needs the folder, where to search, an included list, excluded list and an action list These lists are patterns The action list, decides if the file will be deleted or truncated.
The Collector:
saves the content in a File System
- knows, which content was saved in the File System
- it can connect to many Agents
This application was developed by "our selfs" , but is very unreliable.
Files are deleted without knowing
Allot of spaghetti code and very old
Very little time to repair it and very hard to repair
When I saw the hive2hive, I see many interesting features:
File Synchronization
File Versioning and Conflict Management
File Watchdog / Change Detection (automated, configurable)
Users can use multiple clients (simulatenously)
Multiple users can use the same machine (simultaneously)
I would like to run the application headless, a program that runs in the background.
I have some questions:
- How can I decide which files to synch? I could have a list of files selected by a pattern, for example *.log
How can I send from a source server to a destination server and keep Versioning?
Is it possible to use key files? Most of the examples were with user credentials. I prefer key file
How should I configure it? 4.1. Should I have 2 application? One in the source the other in the destination 4.2. Should I have only one application? Where should it be? Source or Destination
Can I keep the same format, many Agents and few Collectors?
I am the only person, that can implement this application, because of the time I am asking for help.
I would like to hear your opinion and advice. Thank you very much in advance, Best regards,
Luís
