![\"DSCN0067\"](\"https:\/\/aidanfinn.com\/wp-content\/uploads\/2013\/06\/DSCN00675.jpg\" "\\"DSCN0067\\"")
<\/a><\/p>\nHardware<\/span><\/strong><\/p>\n It\u2019s simple and cost effective. Goes from basic to OLTP workloads.<\/p>\n Capabilities Overview<\/span><\/strong><\/p>\n Storage Pools support ReFS<\/p>\n 2 * JBODs.\u00a0 We create a single storage pool to aggregate all 48 disks ( 2 * 24 in this example.\u00a0 We create 1 * 2-way mirror spaces and 1 * parity space.<\/p>\n Data is spread around the disks in the storage pool.\u00a0 They parallelize the rebuild process.<\/p>\n 8 * 3 TB disk test bed.\u00a0 Test the failure of the disk.\u00a0 Can rebuild in 50 minutes, with > 800 MB\/s rebuild throughput.\u00a0 The line is that hot spare is no longer necessary in WS2012 R2.\u00a0 Hmm.\u00a0 Must look into that.<\/p>\n Scale-Out Example<\/span><\/strong><\/p>\n Note: CSV scales out linearly<\/p>\n Match workload characteristics to drives<\/span><\/strong><\/p>\n Can we seamlessly merge these?<\/p>\n Tiered Storage Spaces<\/span><\/strong><\/p>\n A single virtual disk can use the best of both types of disk.\u00a0 High capacity for colder slices of data.\u00a0 High speed for hotter slices of data.<\/p>\n The most compelling ratio appears to be 4 to 12 SSDs in a 60 slot device, with the rest of the disks being HDDs.<\/p>\n In the background, the file system actively measures the activity of file slices<\/span>.\u00a0 Transparently moves hot slices to the SSD tier, and cold slices to the HDD tier.<\/p>\n Tiering (analysis and movement) is done daily.\u00a0 The schedule is configurable (change time, do it more than daily).\u00a0 The slices are 1 MB in size.\u00a0 So tracking watches 1 MB slices, and tiering is done on 1 MB slices.<\/p>\n Administrators can pin entire<\/span> files to specified tiers.\u00a0 Example, move a VDI parent VHDX to the SSD tier.<\/p>\n Write-Back Cache<\/span><\/strong><\/p>\n Per virtual disk, persistent write cache.\u00a0 It smoothens out write bursts to a virtual disk.\u00a0 Uses the SSD capacity of the pool for increased IOPS capacity.\u00a0 Configurable using PowerShell.\u00a0 Great for Hyper-V which needs write-through, instead of battery powered write cache.<\/p>\n PowerShell Demo<\/span><\/strong><\/p>\n Get-PhysicalDisk to list the possible disks to use \u201cCanPool\u201dattribute.<\/p>\n $disks = Get-PhyscalDisks<\/p>\n New-StoragePool \u2026. $disks<\/p>\n Get-StoragePool to see the disks.\u00a0 Look at FriendlyName and MediaType attributes.<\/p>\n $SSD_Tier = New-StorageTier \u2026 list the SSDs<\/p>\n $HDD_Tier = New-StorageTier \u2026 list the HDDs<\/p>\n $vd1 = New-VirtualDisk \u2026. \u2013StorageTiers @($ssd_tier, $hdd_tier) \u2013StorageTierSizes @(150GB, 1.7TB) \u2026.<\/p>\n Now we have a drive with automated scheduled storage tiering.<\/p>\n Pins some files using Set-FileStorgeTier<\/p>\n Optimize-Volume \u2013DriveLetter E \u2013TierOptimize\u00a0 \u2026.. this will force the maintenance task to run and move slices.<\/p>\n Demo: Write-Back Cache<\/span><\/strong><\/p>\n He increases the write workload to the disk. A quick spike and then the SSD takes over.\u00a0 Then increases again and again, and the write-back cache absorbs the spikes.<\/p>\n Question: How many tiers are supported in WS2012 R2?\u00a0 2.\u00a0 But the architecture will allow MSFT to increase this in later releases if required.<\/p>\n Right now, certified clustered storage spaces from:<\/p>\n Takeaways<\/span><\/strong><\/p>\n Q&A<\/span><\/strong><\/p>\n No docs on sizing Write-Back Cache.\u00a0 They want the WBC to be not too large.\u00a0 Up to 10 GB is being recommended right now.\u00a0 You can reconfigure the size of the WBC after the fact \u2026 so monitor it and change as required.<\/p>\n On 15K disks: Expensive and small.\u00a0 Makes sense to consider SSD + 7.5K disks in a storage pool rather than SSD + 15 K in a storage pool.<\/p>\n He can\u2019t say it, but tier 1 manufacturers are scared *hitless of Storage Spaces.\u00a0 I also hear one of them is telling porky pies to people on the Expo floor re the optimization phase of Storage Spaces, e.g. saying it is manual.<\/p>\n Is there support for hot spares?\u00a0 Yes, in WS2012 and R2.\u00a0 Now MSFT saying you should use space capacity in the pool with parallelized repair across all disks in the pool, rather than having a single repair point.<\/p>\n DeFrag is still important for contiguous data access.<\/p>\n If I have a file on the SSD tier, and the tier is full, writes will continue OK on the lower tier.\u00a0 The ReFS integrity stream mechanism can find best placement for a block.\u00a0 This is integrated with tiered storage spaces.<\/p>\n On adding physical disks to the storage space: old data is not moved: instant availability.\u00a0 New writes are sent to the new disks.<\/p>\n A feature called dirty region table protects the storage space against power loss caused corruption.<\/p>\n Should hard drive caches be turned off?\u00a0 For performance: turn it off.\u00a0 For resilience, turn it on.\u00a0 Note, a cluster will bypass the disk cache with write-through.<\/p>\n There is some level of failure prediction.\u00a0 There are PoSH modules for detecting issues, e.g. higher than normal block failure rates, or disks that are slower than similar neighbours.<\/p>\n Ah, the usual question: Can the disks in a storage space span data centers.\u00a0 The members of a storage pool must be connected to all nodes in a SOFS via SAS, which makes that impossible.\u00a0 Instead, have 2 different host\/storage blocks in 2 sites, and use Hyper-V Replica to replicate VMs.<\/p>\n Virtual Disk Deployment Recommendations<\/span><\/strong><\/p>\n When to use Mirror, Parity, or Simple virtual disks in a storage space?<\/p>\n A storage space will automatically repair itself when a drive fails \u2013 and then it becomes resilient again.\u00a0 That\u2019s quick thanks to parallelized repair.<\/p>\n Personal Comment<\/span><\/strong><\/p>\n Love hearing a person talk who clearly knows their stuff and is very clear in their presentation.<\/p>\n —<\/p>\n Holy crap, I have over a mile to walk to get to the next storage session!\u00a0 I have to get out before the Q&A ends.<\/p>\n Speaker: Brian Matthew Start was some metrics achieved stuff.\u00a0 Summary: Lots of IOPS. Hardware It\u2019s simple and cost effective. Goes from basic to OLTP workloads. Capabilities Overview Storage Pools support ReFS 2 * JBODs.\u00a0 We create a single storage pool to aggregate all 48 disks ( 2 * 24 in this example.\u00a0 We create 1 … Continue reading “TechEd NA 2013\u2013Storage Spaces Performance”<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"_uf_show_specific_survey":0,"_uf_disable_surveys":false,"footnotes":""},"categories":[14],"tags":[99,120],"class_list":["post-14808","post","type-post","status-publish","format-standard","hentry","category-eventnotes","tag-storage","tag-windows-server-2012-r2"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t<\/a><\/p>\n<\/a><\/p>\n\n
<\/a>.<\/p>\n\n
<\/a><\/p>\n<\/a><\/p>\n\n
\n
<\/a><\/p>\n