operating system
RHEL OS
At the core of the design, we leverage an already stable OS that has all the bells & whistles - RHEL OS.
Our API code runs on top of ext3/ext4 allowing us to use the code blocks for copy, read, mirror existing traditional data center file systems.
Our File system supports:
- Backward Compatibility - backward compatibility with ext3 and ext2 and Data center filesystems and other SNIA compliant FS'es.
- Allocation Improvements - Storage block allocation efficiently prior to writing them to the disk. This improves read and write performance.
- Timestamp Improvements - Ext4 runs 408 years to the timestamp and supports dates up to May 10, 2446. Timestamps are also measured faster, in nanoseconds.
- Extents reduce the amount of metadata required to map each file’s blocks.
- Multi-block Allocation Improvements - ext4 file system uses multiple allocations that allow the allocation of multiple blocks per call.
- This reduces disk fragmentation.
- Delayed Allocation - The delayed allocation feature allocates blocks only when the file is written to the disk.
- With this feature, cache memory is not filled with unnecessary data, and the performance of the system increases.
- Unlimited Number of Sub-directories - Linux kernel version 2.6.23 supports an unlimited number of sub-directories and a HTree data structure to avoid drops in performance (B-Trees).
- Journal Check-sums - ext4 file system uses the checksum option. This option was introduced to reduce the risk of file corruption. When hardware failure occurs, the blocks become unusable and file corruption occurs. The checksum option constantly checks to see if a block is damaged. This process also improves performance because it shortens the journal time.
- Faster File System Checks - In an ext4 file system, undistributed groups of blocks and inode tables are marked. The time required to run the fsck command is significantly shortened because marked groups are skipped.
- Online De-fragmentation - ext4 file system supports the e4defrag tool that lets users de-fragment individual files or the complete file system.