{"id":964,"date":"2011-11-02T00:02:18","date_gmt":"2011-11-02T06:02:18","guid":{"rendered":"http:\/\/sanuja.com\/blog\/?p=21"},"modified":"2012-09-13T09:13:46","modified_gmt":"2012-09-13T15:13:46","slug":"server-raid-rebuilt","status":"publish","type":"post","link":"https:\/\/sanuja.com\/blog\/server-raid-rebuilt","title":{"rendered":"Server RAID rebuilt&#8230;"},"content":{"rendered":"<p>This is actually the continuation my\u00a0previous\u00a0update on the server. As I noted in that previous post setting up a RAID 5 with six 146 GB drives and two 72 GB drives was very stupid. But I never\u00a0explained\u00a0why this is a stupid idea on my first post. Today we will look at the best way to mix and match different types of harddrives on a RAID system for optimum performance and the\u00a0highest\u00a0storage output.<\/p>\n<p>The the RAID 5 system works by taking all the hard drives with even\u00a0different\u00a0sizes and strip the data on all of them to provide one drive fail backup support for the system. In other words, on a RAID 5 system, you an\u00a0tolerate\u00a0one drive failure\u00a0without loosing data. In this\u00a0method, the RAID controller will take the lowest capacity of all the drives on the array and multiply that lowest\u00a0capacity\u00a0by the number for drives regardless of the\u00a0capacities\u00a0of the larger drives. For example, if we have two 72 GB drives and six 146 GB drives, the RAID controller will see all the eight drives as 72 GB drives. By doing the math for the RAID 5 system: (72 GB * 8)-72GB = 567 GB in total. Considering 146 GB *6 itself is 876 GB, this creates a a lot of wasted space. In addition to this reduced space, the 72 GB HDs are running at 15K RPM while the 146 GB HDs are running at 10 K causing the RAID to operate at the lowest RPM value of 10 K even with the six 15 K drives have the capability to<!--more--> read and write at faster speeds.<\/p>\n<p>Solution when you want to mix and match? Create two (or more) different independent RAID systems! By dismounting the RAID 5 configuration of mixed drives and rebuilding two independent RAIDs, one with the six 147 GB at 10 K drives as a RAID 5 and the other two 72 GB at 15 K as a\u00a0separate\u00a0RAID 1, you can get the\u00a0maximum\u00a0performance and the\u00a0maximum\u00a0storage capacity for the buck. The OS is loaded to the 75 GB drives on RAID 1 giving it faster speeds to boot up and perform operations for the server as well as top class redundancy. This is a lot more than space to load the OS, most of the server roles, and other applications. The slower side with RAID 5\u00a0yielded about 682 GB after \u00a0formatting. The\u00a0theoretical\u00a0yield\u00a0can be calculated as: (146 GB *6)-146GB = 730 GB. This is about\u00a0theoretical\u00a0value of 163 GB more than what I was able to get from all eight hard drives in RAID 5 configuration.<\/p>\n<p>This new system also have another advantage when it comes to RAID drive failures. The RAID 5 and RAID 1 are independent of each other. The RAID 1 has the\u00a0capacity\u00a0to protect data up to one fail drive and the RIAD 5 also offer the same one drive fail\u00a0safety! Also for future upgrade to the system on the RAID 5 side can be done without loosing the OS updates and applications stored in the RAID 1 side of the server.<\/p>\n<p>Why didn&#8217;t I think of this before?<\/p>\n","protected":false},"excerpt":{"rendered":"<p>This is actually the continuation my\u00a0previous\u00a0update on the server. As I noted in that previous post setting up a RAID 5 with six 146 GB drives and two 72 GB drives was very stupid. But I never\u00a0explained\u00a0why this is a stupid idea on my first post. Today we will look at the best way to &hellip; <a href=\"https:\/\/sanuja.com\/blog\/server-raid-rebuilt\" class=\"more-link\">Continue reading <span class=\"screen-reader-text\">Server RAID rebuilt&#8230;<\/span> <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[3],"tags":[18,7,10],"class_list":["post-964","post","type-post","status-publish","format-standard","hentry","category-computer-science","tag-computing","tag-server","tag-tech"],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/posts\/964","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/comments?post=964"}],"version-history":[{"count":0,"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/posts\/964\/revisions"}],"wp:attachment":[{"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/media?parent=964"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/categories?post=964"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sanuja.com\/blog\/wp-json\/wp\/v2\/tags?post=964"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}