Dates and Events: |
OSADL Articles:
2023-11-12 12:00
Open Source License Obligations Checklists even better nowImport the checklists to other tools, create context diffs and merged lists
2022-07-11 12:00
Call for participation in phase #4 of Open Source OPC UA open62541 support projectLetter of Intent fulfills wish list from recent survey
2022-01-13 12:00
Phase #3 of OSADL project on OPC UA PubSub over TSN successfully completedAnother important milestone on the way to interoperable Open Source real-time Ethernet has been reached
2021-02-09 12:00
Open Source OPC UA PubSub over TSN project phase #3 launchedLetter of Intent with call for participation is now available |
Improved Redundancy and Consistency beyond RAID 1
Roland Kammerer, TU Vienna
Bededikt Huber, TU Vienna
The FLOSS community provides a wide variety of file systems which are based on substantially different designs along with independent development. The implementation of safe storage can take advantage of this diversity. RAID level 1 provides an inexpensive technology to mirror data over a set of disks. However, the use of RAID-1 is constrained by limitations of data consistency and integrity checks when data is read from a RAID-1 disk set. For performance reasons, most RAID-1 implementations read chunks of a file from different disks of the set without checking the consistency of a chunk by reading the mirrored data from another disk in the set. This might lead to a scenario where a faulty chunk of a file is read and forwarded to a higher level in the file system hierarchy.
This paper presents a solution to overcome the limitations of RAID-1 by utilizing existing FLOSS technology to implement a simple file system layer that provides sufficient integrity checks when data is read from a set of disks. For write operations the layer acts similar to RAID-1, it mirrors data to different disks. For read operations the presented layer has two different modes: In two-disk mode, data from one disk are compared with a second copy from an additional disk and in case of inconsistency a file system error is reported. In three-disk mode, the layer provides Triple Modular Redundancy with majority voting. As long as two disks agree on the content of a chunk, the file system is operational. Additionally, the layer tries to overwrite the faulty copy with the agreed data, which is sufficient in case of transient faults. With the presented layer, two main advantages are gained in the context of safe storage: First, the data are checked for consistency on read operations. In three-disk mode corrupted data can even be corrected. Second, all disks can have different underlying file systems. This diversity increases the confidence in the whole system, because the trust has not to be placed on a single type of file system.