Recommended Posts

After having an issue with my RAID-5 array recently, that resulted in only 1 lost file, I decided it was time to finally upgrade. Going from 8x 2TB drives in RAID 5, I decided to move to a RAID 6 array, starting with 6x 3TB drives. The RAID 6 will allow me to have better resiliency if a drive dies, but I'm still planning on keeping not only a hot spare around, but a separate 3TB backup drive for critical files that will get backed up to nightly.

So my question is now, how best to partition my new server. I use it for the following:

1.) Hyper-V VM host (~300GB)

  • ?ADDC
  • Exchange 2010
  • Linux Anti spam/virus/phishing filter in front of exchange
  • Apache web server
  • subsonic media streamer
  • various other vms

2.) TV show / Movie repository for streaming to XBMC (~7TB)

3.) SVN Server for code/Electronic schematics/board files/gerbers (~80MB)

and a few other things. The RAID 6 array has a 64KB stripe size, and while it currently has 6x3TB drives, I bought a new case and 4 5-in-3 drive cages to allow me to increase it to 20 3TB (or 4TB in the future) drives.

Should I:

1.) Create one large array, then partition it into separate drives

2.) Create smaller RAID 6 arrays (or possibly RAID-1 arrays for critical stuff)

3.) Create multiple, different type arrays for each specific function above

My requirements would be that if necessary, I should be able to expand a volume if it needs more space (in windows drive manager), and that when space runs out, be able to easily add in and expand the array(s) with new drives. I am familiar with OCE, but im not sure if I can span different arrays on same drives (say a 4TB RAID 6 array, spanning 5x 3TB drives, plus a 4TB RAID-1 array on those same 5 drives). Ideally I'd prefer separate partitions/arrays, so that I can specify different block/strip sizes (SVN stuff is usually small, so say 32 or 64KB for that, whereas TV and movies are usually several gigabytes, so maybe a 512 or 1024KB strip for those).

Anyone have any recommendations?

so FWIW,I decided to do the following: I bought a SAS expander and several 5-in-3 cages, and will do the following config:

1.) 4x3TB in RAID-5 + 1 x 3TB Hotspare for movies and TV shows

2.) use 2 of my old 2TB drives to do a RAID-1 array for my SVN, photography, website, etc.

3.) grab 4x1TB, or possibly 4 of the 2TB drives I had and do a 0+1 for my Hyper-V VMs.

This way I can always plug in extra drives and either expand, migrate, etc, and I've got extra slots so I can put in other backups.

For the videos use refs or ntfs with 64k clusters

Yeah, originally i was trying to figure out how to best strip a single raid array for multiple uses, but I think it's better to do multiple arrays on the controller, then use something like:

512KB/1MB strip for the video content (64KB REFS)

32KB-64KB strip for SVN/photography(NTFS)

128-256KB strip for Hyper-V vms (64KB REFS system)

This topic is now closed to further replies.
  • Posts

    • Digisecret 2.1.431 Pro and Wzipse 4.0. Both are encrypted self-extracting archives.
    • Google reshuffles its AI coding team as it struggles to catch Anthropic by Karthik Mudaliar Google is already reorganizing the AI coding “strike team” it created roughly two months ago, as it attempts to find ways to close the gap with Anthropic in one of generative AI’s most commercially important areas. According to The Information, Google DeepMind is expanding the team’s focus to include “midtraining,” rather than concentrating only on coding tools and agents. Midtraining takes place after a model’s broad initial training but before the final stages that prepare it to follow instructions and perform specific tasks. In simple terms, it gives developers another opportunity to expose a model to carefully selected data before it is polished for release. That could help Google improve Gemini’s underlying coding abilities instead of relying only on better prompts, interfaces, or post-training. Previous research has found that midtraining can be particularly effective for code and mathematics, where models must move from general language knowledge to more structured tasks. Google reportedly created the original strike team in April. It was led by Google DeepMind research engineer Sebastian Borgeaud, who previously worked on model pretraining, and focused on complex, long-running programming jobs. Google co-founder Sergey Brin and Google DeepMind chief technology officer Koray Kavukcuoglu were also reportedly involved in the effort. DeepMind researchers were said to believe that Anthropic’s coding tools were outperforming Google’s Gemini models, prompting the company to give the project more attention. Anthropic has made coding a central part of its AI strategy through Claude Code and its Claude model family. The company has continued improving that area, with Claude Opus 4.8 offering upgrades for coding and other agentic tasks, along with the now-unavailable Mythos and Fable models. The reshuffle also comes at a time when Google faces increased competition for AI researchers. Gemini co-lead Noam Shazeer recently announced that he was leaving Google for OpenAI, while two other researchers who contributed to Gemini and DeepMind projects are reportedly preparing to join Anthropic. It remains unclear whether the reorganized team will produce a new public Gemini model or developer product. No release date, team size, or specific performance target has been disclosed. Source: The Information
  • Recent Achievements

    • First Post
      kinowa earned a badge
      First Post
    • Rookie
      krychek57 went up a rank
      Rookie
    • Grand Master
      Jaybonaut went up a rank
      Grand Master
    • One Year In
      Philsl earned a badge
      One Year In
    • Dedicated
      Scoobystu earned a badge
      Dedicated
  • Popular Contributors

    1. 1
      +primortal
      413
    2. 2
      +Edouard
      168
    3. 3
      PsYcHoKiLLa
      132
    4. 4
      Xenon
      73
    5. 5
      Michael Scrip
      73
  • Tell a friend

    Love Neowin? Tell a friend!