Parmos Designgoals:

Costeffecitve high load, high availabilty data processing clusters.

Base concepts:

Utilising cost effective distributed Intel computers and AT-Bus base harddrive under LINUX. Ensuring the optimal use of this configuration by specialised software, optimized to balance the lower overall availbilty and minimise the administrational overhead.

Methods:

Centalised Administration of computer nodes. Using "Plug and Play" concepts to add or administer new nodes. Centralised controll to distribute data to more than one node, to maximise performance and availabilty. Spliting up processes normaly done within the database to minize the hardware requirements:
  1. Transactionsave storage of events.
  2. Centralized administration of all log files on all logger/storage nodes.
  3. Translating the storage structure into an database like model and usage of asynchrones SQL:1999 compatible command language for data preprocessing.
  4. Uploading the derived data into an online database.

Availabilty:

Parmos is currently within the preparation and controll phase. Not all parts of the parmos system have currently reached a level where they could be sold out of the box. If you have concreat need for information or thinking of a parmos installation please mail to info@parmos.org. We would be pleased to support you with additional information.