Home › Studies › AZ-305 › Data Storage › Managed Disks

Managed Disks

Ultra · Premium SSD v2 · Premium SSD · Standard SSD · Standard HDD

Practice questions for this topic →

Disk types

TypeMax IOPSMax throughputMax sizeOS disk?Use for
Ultra Disk400,00010,000 MB/s65,536 GiBNoSAP HANA, top-tier databases; change performance without restarting the VM
Premium SSD v280,0002,000 MB/s65,536 GiBNoPerformance-sensitive production; tune IOPS and throughput independently of size
Premium SSD20,000900 MB/s32,767 GiBYesProduction VMs and OS disks
Standard SSD6,000750 MB/s32,767 GiBYesWeb servers, light apps, dev/test
Standard HDD2,000 (3,000 with performance plus)500 MB/s32,767 GiBYes (retiring Sep 8, 2028)Backup, non-critical

Facts to know

Exam Tips

Tick each tip once you can answer it without looking.

“SAP HANA / extreme IOPS, adjust performance live” → Ultra Disk.

“High IOPS database data disk, tune performance independently of size, lower cost than Ultra” → Premium SSD v2.

“Need high-performance OS disk” → Premium SSD. Trap: Ultra and Premium SSD v2 can't be OS disks.

“Dev/test VMs, cheapest reasonable performance” → Standard SSD.

“Disk must survive a zone failure while attached” → ZRS managed disk.

“Failover cluster needs a shared disk” → Azure shared disks.

“Standard HDD OS disks” → Retiring Sep 8, 2028, so plan Standard SSD.

“Compare maximum IOPS” → Ultra 400,000; Premium SSD v2 80,000; Premium SSD 20,000; Standard SSD 6,000; Standard HDD 2,000 (3,000 with performance plus). Trap: Only Premium SSD and below can be OS disks.

“Default Ultra Disk capacity per region per subscription” → 100 TiB (request a quota increase for more). Trap: Ultra Disks are data disks and can only be created empty.

“Disk types that reach 64 TiB” → Ultra Disk and Premium SSD v2 (65,536 GiB). Trap: Premium SSD, Standard SSD and Standard HDD top out at 32,767 GiB.

“Ultra Disk size and performance” → 4 GiB to 64 TiB; IOPS cap grows with size, reaching 400,000 from 512 GiB. Trap: Performance can be changed without restarting the VM.

“Durability and availability of managed disks” → Designed for 99.999% availability; data replicated three times; LRS disks at least 11 nines of durability, ZRS disks at least 12 nines. Trap: Durability isn't the same measure as a VM's uptime SLA.

“Keep disk data accessible through a zone outage” → ZRS managed disks; also use ZRS when sharing disks between VMs. Trap: A shared disk on LRS is a single point of failure.

Sources

← Azure FilesData Integration →