At its heart, mdraid:<\/p>\n\n\n\n
\u00b7 Lives primarily in the Linux kernel as the md<\/em> (Multiple Device) driver.<\/p>\n\n\n\n \u00b7 Uses on\u2011disk superblock metadata to describe array membership and layout.<\/p>\n\n\n\n \u00b7 Exposes assembled arrays as standard block devices that can be partitioned, formatted, LVM\u2019ed, encrypted, or used directly by applications.<\/p>\n\n\n\n In other words: mdraid is the engine; mdadm is the toolkit and dashboard.<\/p>\n\n\n\n A simplified vertical stack (bottom \u2192 top):<\/p>\n\n\n\n [Physical \/ Virtual Drives] Because md devices appear as regular block devices, you can build rich storage stacks: encrypt then RAID, RAID then encrypt, layer LVM for flexible volume management, or present md devices to virtualization hosts.<\/p>\n\n\n\n Remember:<\/strong> You manage arrays with mdadm; the arrays themselves are provided by the mdraid kernel layer.<\/p>\n<\/blockquote>\n\n\n\n mdraid supports a broad set of personalities (RAID implementations). Availability may vary slightly by kernel version, but common ones include:<\/p>\n\n\n\n \u00b7 linear<\/strong> \u2014 Concatenate devices end\u2011to\u2011end; no redundancy.<\/p>\n\n\n\n \u00b7 RAID 0 (striping)<\/strong> \u2014 Performance & aggregate capacity; zero redundancy.<\/p>\n\n\n\n \u00b7 RAID 1 (mirroring)<\/strong> \u2014 Redundancy via copies; read parallelism.<\/p>\n\n\n\n \u00b7 RAID 4<\/strong> \u2014 Dedicated parity disk; rarely used.<\/p>\n\n\n\n \u00b7 RAID 5<\/strong> \u2014 Distributed parity across members; 1\u2011disk fault tolerance.<\/p>\n\n\n\n \u00b7 RAID 6<\/strong> \u2014 Dual distributed parity; 2\u2011disk fault tolerance.<\/p>\n\n\n\n \u00b7 RAID 10<\/strong> \u2014 Striped mirrors (mdraid implements as an n\u2011way layout over mirrored sets; flexible). Good mix of speed + redundancy.<\/p>\n\n\n\n \u00b7 RAID 1E \/ RAID 10 variants<\/strong> \u2014 Extended \/ asymmetric mirror\u2011stripe layouts for odd numbers of drives.<\/p>\n\n\n\n \u00b7 Multipath<\/strong> \u2014 md can be used (less common today) to provide failover across multiple I\/O paths.<\/p>\n\n\n\n \u00b7 Faulty<\/strong> \u2014 Testing personality that injects failures.<\/p>\n\n\n\n Superblock Metadata<\/strong> \u2014 Small headers stored on member devices that describe array UUID, RAID level, layout, chunk size, role of each device, and state.<\/p>\n\n\n\n md Personalities<\/strong> \u2014 RAID level implementations registered with the kernel\u2019s md subsystem.<\/p>\n\n\n\n Bitmaps<\/strong> \u2014 Optional on\u2011disk or in\u2011memory bitmaps track which stripes are dirty, dramatically shortening resync\/recovery after unclean shutdowns.<\/p>\n\n\n\n Reshape Engine<\/strong> \u2014 Allows changing array geometry (add\/remove devices, change RAID level in some cases) online in many scenarios. Performance impact can be significant; plan maintenance windows.<\/p>\n\n\n\n mdmon (External Metadata Monitor)<\/strong> \u2014 Required for some external metadata formats (e.g., IMSM\/Intel Matrix) to keep metadata in sync.<\/p>\n\n\n\n Sysfs Controls<\/strong> \u2014 Tunables under \/sys\/block\/mdX\/md\/ expose runtime parameters (stripe cache, sync speed limits, write\u2011mostly flags, etc.).<\/p>\n\n\n\n \u00b7 Homelab NAS or DIY storage server using commodity disks.<\/p>\n\n\n\n \u00b7 Bootable mirrors (RAID1) for root filesystem resiliency.<\/p>\n\n\n\n \u00b7 RAID10 for database or virtualization workloads needing strong random I\/O + redundancy.<\/p>\n\n\n\n \u00b7 RAID5\/6 for bulk capacity with parity protection (backup targets, media libraries \u2014 though see risk discussion below on rebuild windows and URE rates).<\/p>\n\n\n\n \u00b7 Aggregating NVMe drives when hardware RAID isn\u2019t desired or available.<\/p>\n\n\n\n \u00b7 Cloud or virtual environments where virtual disks need software\u2011defined redundancy across failure domains.<\/p>\n\n\n\n Before typing mdadm \u2014 create, answer these:<\/p>\n\n\n\n 1. Workload Profile<\/strong> 2. Redundancy vs Capacity<\/strong> 3. Media Type<\/strong> 4. Growth Expectations<\/strong> 5. Boot Requirements<\/strong> 6. Monitoring & Alerting Plan<\/strong> 7. Backup Strategy<\/strong> More & posted from: https:\/\/medium.com\/@pltnvs\/the-definitive-guide-to-mdraid-mdadm-and-linux-software-raid-fbb561c21878<\/a><\/p>\n","protected":false},"excerpt":{"rendered":" What Is mdraid? Core Concepts mdraid (often shortened to MD RAID or simply md) is the Linux kernel\u2019s built\u2011in software RAID framework. It aggregates multiple block devices (drives, partitions, loopbacks, NVMe namespaces, etc.) into a single logical block device (\/dev\/mdX) that can deliver improved performance, redundancy, capacity aggregation \u2014 or some combination \u2014 depending on the selected RAID level.… Read More »The Definitive Guide to mdraid, mdadm, and Linux Software RAID|Medium<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":18551,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"neve_meta_sidebar":"","neve_meta_container":"","neve_meta_enable_content_width":"","neve_meta_content_width":0,"neve_meta_title_alignment":"","neve_meta_author_avatar":"","neve_post_elements_order":"","neve_meta_disable_header":"","neve_meta_disable_footer":"","neve_meta_disable_title":"","footnotes":""},"categories":[384,17],"tags":[],"class_list":["post-18550","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-computer-science","category-linux"],"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/posts\/18550","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/comments?post=18550"}],"version-history":[{"count":1,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/posts\/18550\/revisions"}],"predecessor-version":[{"id":18552,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/posts\/18550\/revisions\/18552"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/media\/18551"}],"wp:attachment":[{"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/media?parent=18550"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/categories?post=18550"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/roglacup.com\/klaus62\/wp-json\/wp\/v2\/tags?post=18550"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}How mdraid Fits in the Linux Storage Stack<\/h2>\n\n\n\n
SATA \/ SAS \/ NVMe \/ iSCSI LUNs \/ VMDKs \/ Cloud Block Vols
\u2193
[mdraid Kernel Layer] \u2190 assembled via mdadm
RAID0 | RAID1 | RAID4\/5\/6 | RAID10 | RAID1E | linear | multipath | etc.
\u2193
[Optional Layers]
LUKS dm-crypt | LVM PV\/VG\/LV (incl. LVM-thin) | Filesystems (ext4, XFS, btrfs, ZFS-on-Linux as zvol consumer, etc.)
\u2193
[Applications \/ Containers \/ VMs]<\/p>\n\n\n\nmdadm vs mdraid: Terminology Clarified<\/h2>\n\n\n\n
![\"\"\/](\"https:\/\/miro.medium.com\/v2\/resize:fit:1050\/1*Rnd6ZBBZCfxvwBLimgS9cQ.png\")
<\/figure>\n\n\n\n\n
Supported RAID Levels & Modes<\/h2>\n\n\n\n
Key Architecture Elements<\/h2>\n\n\n\n
Common Use Cases<\/h2>\n\n\n\n
Planning an mdraid Deployment<\/h2>\n\n\n\n
Random vs sequential; read\u2011heavy vs write\u2011heavy; mixed DB + VM vs cold archive.<\/p>\n\n\n\n
How many failures must you tolerate? RAID1\/10 vs RAID5\/6 trade\u2011offs.<\/p>\n\n\n\n
HDD, SSD, NVMe, or mixed? Parity RAID on SSD behaves differently (write amplification, TRIM behavior, queue depth scaling).<\/p>\n\n\n\n
Will you add drives later? Consider RAID10 or start with larger drive count in RAID6; know reshape implications.<\/p>\n\n\n\n
If booting from md, choose metadata format (0.90 or 1.0) that places superblock where firmware\/bootloader expects.<\/p>\n\n\n\n
Email alerts, systemd timers, mdadm \u2014 monitor, integration with Prometheus\/node_exporter, etc.<\/p>\n\n\n\n
RAID \u2260 backup. Always maintain off\u2011array copies of critical data.<\/p>\n\n\n\n