OpenBSD CVS RAID Array Failing, Needs Replacement

Sam writes "The OpenBSD cvs server has a failing RAID array. Users of the projects on that array: OpenBSD, OpenSSH, OpenBGPD, OpenNTPD, and the upcoming OpenCVS are all invited to contribute towards the $12,500 cost of a suitably high-spec replacement. OpenBSD Journal article, and original request (thread)."

Comments

Re:$12,500... what is this, MAC hardware?

[frost22]

Marco may be as brilliant as God the Almighty ... but that doesn't change facts. SCSI drives and SCSI based arrays are currently in the process of becoming a high end niche product.

Especially against good caching SATA arrays you will have a hard time to even state a convincing performance case, not to mention one based on costs.

Im not arguing those are better. I argue he doesn't need those extra 10 or what percent faster seek times. Over all system design is more than knowing disk to the last atoms. If Marco is responsible fr this design, he confused his desire to have superior Hard drives with his project's desire to maximise a price-performance ratio under a given minimum performance requirement.

Re:Trollfeed:

[frost22]

And you get substantially less throughput too.

Dream on. Throughput isn't even a topic there any more, hasn't been for years, actually. IDE drives have been beating SCSI hands down on this for years now.

The real stated advantage of SCSI drives is seek time. Seek time is the holy grail when it comes to database applications. The key is not to shovel so and so many gigs around - the key is running around super fast and picking a few bytes here and there - thats essentially what a database does.

But (and thats a Big Butt) Theo and his merry men don't need that sort of performance here. They just have some CVS server (which, database wise, is probably rather small, just a few gigs). Most of this will be eventually be cached in RAM anyway.