QStar, a confidential leader with Archive Manager

Specialist in unstructured data archiving, QStar Technologies, founded in 1987, still is very confidential even if the company secured more than 19,000 installations since its inception. This level of adoption is unique in the domain confirming that the firm is profitable operating at its own pace.

QStar CEO, Riccardo Finotti, CTO, Max Finotti, and SVP Sales and Marketing, Dave Thomson, joined yesterday The IT Press Tour in Rome for a company and product update. They even launched a product I'll cover in a few days.

The team has developed a comprehensive product line with Archive Manager, the heart of the solutions set, Network Migrator, Archive Replicator and Object Storage Manager.

Archive Manager (AM) operates as a the archiver controller for data managing storage units dedicated to this behavior and the lifecycle of submitted files. Exposing multiple access methods like NFS, SMB or S3, the product could be seen as a gateway in front of tape libraries or cloud. Implementing a S3-to-Tape model, AM supports also any backup tool via S3 or file sharing protocols such as Cohesity, HYCU or Rubrik. AM introduced the notion of disk cache with a special file system designed by the company and also a special format on tapes. The AM server must have 32GB of RAM minimum for each volume "attached" with the destination storage units. A minimum of 1TB of space is needed for that cache and it appears to be pretty small especially today with 24TB HDD or even larger. Same remark as this size must also consider the largest file size submitted to AM and of course disk array with RAID, erasure coding with today NVMe could be the right choice. As the performance of the cache is key for the global quality of service, considering flash for the cache makes really sense.

Archive Replicator is in fact the same product with the capability to remote copy data sets synchronously to 4 other storage units. Network Migrator is a HSM working with agent, API or pull mode and replacing migrated files by stubs.

But it appears that for larger configurations, a single node for archive is not enough inviting users to multiple this single node configuration. We also anticipate some cluster mode for Archive Manager able to aggregate all nodes performance and deliver this needed performance boost to address large deployments with high volumes of data and huge number of files. More news soon.