1. © 2011 IBM Corporation Sizing Guidelines Jana Jamsek ATS Europe

2. © 2011 IBM Corporation IBM Storage Systems IBM i performance data  Good to use IBM i performance data even before modelling with Disk Magic, to apply sizing guidelines  Needed reports for sizing external storage –System report / Disk utilization & Storage pool utilizaton –Resource report / Disk utilization –Component report / Disk activity  Desired way of collecting: –Collect performance data during 24 hours on 3 consecutive days and during heavy End-of-month job –Colleciton interval 5 minutes  Insert the reports to Disk Magic to obtain data in Excell spreadsheet

3. © 2011 IBM Corporation IBM Storage Systems Example: Spreadsheet by Disk Magic

4. © 2011 IBM Corporation IBM Storage Systems Component report: present cache hits

5. © 2011 IBM Corporation IBM Storage Systems Sizing the disk drives in external storage  DS8800 Recommended maximal disk utilization: 60% –15K RPM SAS disk drives –10 K RPM SAS disk drives –SSD  DS5000 Recommended maximal disk utilization: 45% – 15K RPM disk drives – 10 K RPM disk drives – SSD  XIV –Data modules  Storwize V7000 Recommended maximal disk utilization: 45% –15K RPM SAS disk drives – 10 K RPM SAS disk drives – SSD

6. © 2011 IBM Corporation IBM Storage Systems Guidelines for RAID level  RAID-10 provides better resiliency  RAID-10 provides gererally better performance: –RAID-5 results in 4 disk operations per write – higher penalty –RAID-10 results in 2 disk operations per write – lower penalty  RAID-10 requires more capacity  In DS8000 use RAID-10 when: – There are many random writes – Write cache efficiency is low – Huge workload  In Midrange storage and Storwize V7000 we recommend to use RAID-10

7. © 2011 IBM Corporation IBM Storage Systems DS8800: Number of ranks  Detailed calculation for maximal IO/sec on Raid-5 rank  ( reads/sec – read cache hits % ) + 4 * (writes/sec – write cache efficiency) = disk operations/sec (on rank)  One 6+P 15 K rpm rank can handle max 2047 disk accesses/sec, at recommended 60% utilization: 1228 disk ops/sec  Divide current disk accesses/sec by 1228  Example: 261 reads/sec, 1704 writes/sec, 45% read cache hits, 24 % write efficiency: (261-117) + 4 *(1704- 409) = 5324, 5324 / 1228 = 4 - 5 ranks  Recommended: 4 ranks  Calculation is based on performance measurements in Storage development and recommended % disk util.

8. © 2011 IBM Corporation IBM Storage Systems DS8800: Number of ranks - continue  Estimate % read cache hit and % write cache efficiency from present cache hits on internal disk  Rough estimation by best practise: –If % cache hits is below 50% estimate the same percentage on external storage – If % cache hits is above 50% estimate half of this % on external storage  If cache hits are not known or you are in doubt, use Disk Magic default estimation: 20% read cache hit, 30% write cache efficiency

9. © 2011 IBM Corporation IBM Storage Systems DS8800: Number of ranks - continue  Quick calculation based on shown detailed calcuation  Example 9800 IO/sec with Read/Write ratio 50/50 need 9800 / 982 =app10 * RAID-10 ranks of 15 K rpm disk drives, connected with IOP- less adapters  The table can be found in the Redbook IBM System Storage DS8000:Host Attachment and Interoperability, SG24-8887-00 Assumed cache hits: 20% read hit 30% write efficiency

10. © 2011 IBM Corporation IBM Storage Systems DS5000/4000/3000:Number of disk drives  Detailed Calculation for maximal IO/sec on disk in Raid-10  ( reads/sec – read cache hits % ) + 2 * (writes/sec – write cache efficiency) = disk operations/sec (on disk)  Quick calculation  IO/sec per DDM: 70% Read50% Read 15 K RPM disk drive RAID-1 or RAID-10 8274 RAID-55845 10 K RPM disk drive RAID-1 or RAID-10 5549 RAID-53930 Example: 7000 IO/sec with read/write ratio 70 / 30 needs 7000 / 82 =app 85 * 15 K RPM disk drives in RAID-10

11. © 2011 IBM Corporation IBM Storage Systems Storwize V7000: Number of disk drives  Quick calculation: 70% Read50% Read 15 K RPM disk drive RAID-1 or RAID-10 138122 RAID-5 9675 10 K RPM disk drive RAID-1 or RAID-10 9282 RAID-5 6450 Example: 7000 IO/sec with read/write ratio 70 / 30 needs 7000 / 138 =app 50 * 15 K RPM disk drives in RAID-10

12. © 2011 IBM Corporation IBM Storage Systems Number of DDMs, connected with VIOS and SVC  The sizing guidelines and calculations for DDMs in Storage systems connected with VIOS or VIOS_NPIV don’t change  The sizing guidelines and calculations for DDMs in Storage systems connected with SVC and VIOS don’t change

13. © 2011 IBM Corporation IBM Storage Systems Sizing for big blocksizes (transfer sizes)  Big blocksize: 64 KB and above  Add about 25% disk arms for big blocksizes  The shown guidelines assume small blocksize (about 12 KB)  Peak in IO usually experiences small blocksizes  Peak in blocksizes has typically low IO/sec  So usually we size for peak in IO/sec and don’t update with additional 25%

14. © 2011 IBM Corporation IBM Storage Systems Number and size of LUNs  With given disk capacity: The bigger the number of LUNs the smaller the size  Sizing guidelines for the number, or for the size  To obtain the number of LUNs you may use WLE ( number of disk drives)  Considerations for very big number of LUNs: –Many physical adapters are needed for natively connected storage –Difficult to manage and troubleshoot with big number of virtual adapters in VIOS

15. © 2011 IBM Corporation IBM Storage Systems  DS8000 Guideline by best practise: – 2 * size of LUN = 1 * size of DDM, or – 4 * size of LUN = 1 * size of DDM – Presently 70 GB or 140 GB LUNs are mostly used  DS5000 Guideline: –Big LUNs enable better seek time on disk drives –Small LUNs – big number of LUNs enable more concurrent IO to disk space –Compromise: 70 GB or 140 GB LUNs  DS5000 best practise: –146 GB physical disks –Make RAID-1 arrays of tow physical disks and create one logical drive per RAID-1 array. –Recommended segment size 128 KB or 64 KB –Create one LUN per array – If the number of LUNs is limited to 16 (for example, connecting to IBM i on BladeCenter) you may want to make a RAID-10 array of four physical disks and create one LUN per array. Number and size of LUNs -continue

16. © 2011 IBM Corporation IBM Storage Systems Number and size of LUNs -continue  XIV - best pracise for the size of LUNs:  Measurements in Mainz: –CPW 96000 users, 2 concurrent runs in different LPARs –15-module XIV Gen 3 Transaction resp. Time (sec) Disk service time (ms) Latency in XIV% Cache hits in XIV 42 * 154 GB LUNs 6.64.6480 6 * 1TB LUNs19.38.2760 70 GB LUNs were not tested Recommendation: about 140GB LUNs, or 70GB LUNs

17. © 2011 IBM Corporation IBM Storage Systems Number and size of LUNs -continue  Storwize V7000, SVC  Presently we recommend about 140 GB LUNs  This recommendation is based on best practise by other midrance storage systems  Recommended to craete vdisks in Striped mode ( default )  Recommended extent size 256 MB ( default )

18. © 2011 IBM Corporation IBM Storage Systems Guidelines for different types of connection  The listed guidelines for a particular storage system apply to all cases (when applicable): – Native connection Sizing for physical FC adapters applies to natively connected storage – Connectioned with VIOS vscsi – Connection with VIOS_NPIV – Connection via SVC The size of LUNs applies to SVC vdisks

19. © 2011 IBM Corporation IBM Storage Systems Sizing FC adapters in IBM i – by IO/sec #5774 or #5749#5735 IO/sec at 70% utilization 10500 per port12250 per port GB per port in adapter 28003266  Example: for one port in #5735 we roecmmend 3266 / 70 = 46 * 70 GB LUNs  Example: For 2 ports in Multipath we recommend 2 * 46 = 92 -> 64 * 70 GB LUNs in Multipath Assumed: Access Density = 1.5, For 2 path: multiply the capacity by 2

20. © 2011 IBM Corporation IBM Storage Systems Throughput of IOP-less adapters #5774 or #5749#5735 Max sequential throughput 310 MB/sec per port400 MB/sec Avg of max sequential throughput for 4KB, 256KB raeds, writes 216 MB/sec per port Avg of min sequential throughput for 4KB, 256KB raeds, writes 132 MB/sec per port Max transaction workload throughput 250 MB/sec per port Transaction workload throughput at 70% utilization 175 MB/sec per port Avg of max sequential throughput for 4KB, 256KB raeds, writes 382 MB/sec per 2 ports Avg of min sequential throughput for 4KB, 256KB raeds, writes 208 MB/sec per 2 ports

21. © 2011 IBM Corporation IBM Storage Systems Adapters on HSL loops

22. © 2011 IBM Corporation IBM Storage Systems Sharing or dedicating ranks  Sharing ranks among multiple IBM i LPARs – Enables better usage of resources in external storage – On the other hand the performance of an LPAR might be influenced by workloads in other LPARs  Dedicating ranks to each LPAR – Enables stable performance (no influence from other systems) – Resources are not as well utilized as with shared ranks  Best practise: – Dedicate ranks to big and/or important systems – Share ranks among medium and small LPARs

23. © 2011 IBM Corporation IBM Storage Systems Guidelines for cache size in external storage  Modelling with Disk Magic  Rough guidelines for DS8800 –10 to 20 TB capacity: 64GB cache – 20 to 50 TB: 128 GB cache – > 50 TB: 256 GB cache

24. © 2011 IBM Corporation IBM Storage Systems Number of HA cards in DS8800  Rules of thumb for HAs in DS8800  About 4 to 8 * ports in IBM i per one HA card in DS8800  For high performance: The number of HA cards should be the same or bigger than the number of device adapters in DS8800  At least one HA card per IO enclosure in DS8800

25. © 2011 IBM Corporation IBM Storage Systems Sizing for VIOS

26. © 2011 IBM Corporation IBM Storage Systems Sizing IASP  Very rough guideline: about 80% of IO will be done to IASP  System report – Resource utilization; IO to database

27. © 2011 IBM Corporation IBM Storage Systems Sizing for SSD

28. © 2011 IBM Corporation IBM Storage Systems Sizing for Metro Mirror or Global Mirror links