i want to install alfresco. But befor i'll do this, i want to ask you few questions:
The architecture of the system should be as follows: 4 Workstations —- LAN —— Server with alfresco —- LAN—– NAS So the alfresco DMS should store the data on the NAS.
Questions: 1. Is the architecture ok? 2. What happends, if the server chrashes: Is it possible to read the data also without the server? –> Fail-safe?
You could read the files, but the are all stored as year/month/day/hour/minute/uuid.bin so you wold have a lot of digging to do. Instead, read http://wiki.alfresco.com/wiki/Backup_and_Restore for proper recovery strategies.
Also, your lucene index is recommended to store in local disk, since it affects performance if read/writes are slow. Since this is derived data, it can always be recreated, it is however recommended that you backup your lucene index (or actually backup-lucene-index) since it will make for a faster recovery it you need to restore.
How many space is required for the lucence index? Or, to put it in other way: Is the lucence index yust a little file beside the .bin or is it the big .bin file (in which the files are stored) which size depends on amount of the stored files?
Have a look in alf_data, there you will find the lucene-indexes folder (wiht many subfolders). You will not find a backup-lucene-indexes folder until you have run Alfresco overnight. It is created 3 am for you to use when you backup. As for the size, it all depends on what you have in your repository. Only Images (that do not output text to be indexed) -> smaller index. I havent actually looked for this in a large live repo, but something around 10-20% of you content size. Maybe someone can fill in here and give an actual example.