In current data center environments where on an average the storage requirements doubles every 12 to 18 months, Storage Optimization becomes a very important function and technique, that needs to be incorporated into the storage architecture and design ahead of time and then followed up constantly as part of the Capacity Planning process on regular basis. When we say Capacity Planning, we include satisfying storage performance requirements (performance SLA\u2019s) besides making sure that enough storage capacity exists in terms of gigabytes to manage storage operations.<\/p>\n
Both Storage Capacity and Storage Performance are key aspects of any Storage Optimization methodology. We want to optimize our data center storage utilizations and hardware utilizations to have the best cost optimized solution that meets the requirements on daily basis\u00a0 and provides enough headroom to meet any SLA\u2019s that are required by the end-user applications.<\/p>\n
For a near holistic Storage Optimization strategy we need to make sure that the various layers from storage arrays to host file systems are managed in an optimized fashion on the minimum. The database layer should be included (e.g. DB files or raw DB devices) to be more complete in the approach. At a high level the storage optimization can be categorized at the storage capacity level and at the storage performance level. Keeping this in mind, a detailed 10-step Storage Optimization methodology can be as follows:<\/p>\n
Capacity Based Optimization \u2013 Storage Array and Host Level<\/b><\/p>\n
STEP 1<\/b> \u2013 Baseline Capacity allocations for each host at the storage array.<\/p>\n
STEP 2<\/b> \u2013 Identify any hosts that are offline (not logged into SAN) \u2013 This is a low hanging fruit in terms of reclaiming storage – Opportunity for storage reclamation.<\/i><\/p>\n
STEP 3<\/b> – Compare storage allocated at the array level\u00a0 for the host (at WWN level) versus\u00a0 storage actually configured at the host OS Level \u2013 Opportunity for storage reclamation if the host level shows less.<\/i><\/p>\n
STEP 4<\/b> – Calculate overall File system utilizations at host level (e.g. running \u201cdf\u201d command on unix systems)\u2013 Opportunity for storage reclamation if utilizations are low. <\/i><\/p>\n
STEP 5<\/b> – Calculate overall Database File utilizations at database level \u2013 Opportunity for storage reclamation if utilizations are low at the DB file level.<\/i><\/p>\n
Performance Based Optimization \u2013 Host IO Profile, but at Storage Array Level<\/b><\/p>\n
STEP 6<\/b> – Baseline Performance Profile for each host at the storage array level for a\u00a0 24 hour peak day. This can be done by aggregating IOPS for all devices masked to the host WWN. Calculate Peak IOPS and 95th<\/sup> Percentile for best practice planning.<\/p>\n STEP 7<\/b> – Match IO Profiles for each host to a currently defined Performance Tier. The SLA for this IO profile could be matched using a lower tier. Many hosts may be able to go to a lower tier, without impacting SLA for application performance \u2013 Opportunity for lowering cost in terms of $\/GB for cost optimization.<\/i><\/p>\n Performance Based Optimization \u2013 Storage Array Level<\/b><\/p>\n STEP 8<\/b> – Identify storage arrays that have low IOPS and MBPS load as compared to the hardware configuration \u2013 opportunity for consolidation.<\/i><\/p>\n STEP 9<\/b> – Create a best-fit consolidation plan based on capacity, performance (IOPS & MBPS) and configuration for all storage arrays not being fully utilized from hardware resource perspective.<\/p>\n STEP 10<\/b> – Migrate hosts at array level based on detailed consolidation plan making sure best-fit analysis is also done for storage array front-end load planning.\u00a0 This leads to lower TCO at the array level by defining optimized storage pools and having the right number of front-end and back-end adapters.<\/p>\n The nice thing about this methodology is that the one can start from step 1 and go as deep as needed based on tools and resource availability. For example, one can only do step 1 and step 2 for storage reclamation as these are the least resource intensive. The OEM tools can be leveraged as provided by the storage vendors to do these but it will take time to get the proper planning done. Home grown tools can also be developed to perform this optimization.<\/p>\n Tools like Perfonics\u2122 developed by Interscape Technologies Inc. and pro-active services like Performance As A Service\u2122 can be leveraged by contacting Interscape Technologies Inc. These services provide a 3T approach to the any type of project by bundling the Technician (SME), Tool (Perfonics\u2122) and the Team (R&D) as part of the project.<\/p>\n Alok Jain<\/p>\n","protected":false},"excerpt":{"rendered":" The 10-Step Storage Optimization Methodology In current data center environments where on an average the storage requirements doubles every 12 to 18 months, Storage Optimization becomes a very important function and technique, that needs to be incorporated into the storage architecture and design ahead of time and then followed up constantly as part of the\u2026 Read More »<\/a><\/span><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[85],"tags":[80,94,53,86],"class_list":["post-316","post","type-post","status-publish","format-standard","hentry","category-optimization","tag-capacity-planning","tag-optimization","tag-storage","tag-storage-performance"],"_links":{"self":[{"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/posts\/316","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/comments?post=316"}],"version-history":[{"count":15,"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/posts\/316\/revisions"}],"predecessor-version":[{"id":472,"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/posts\/316\/revisions\/472"}],"wp:attachment":[{"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/media?parent=316"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/categories?post=316"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.interscape.io\/blog\/wp-json\/wp\/v2\/tags?post=316"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}