IBM DS3300 ESRP Storage Solution Guide

IBM System Storage DS3300 Storage System 1000 Mailbox

Clustered Continuous Replication Microsoft Exchange 2007 Storage

Solution

Tested with: ESRP Storage Version 2.0 Tested Date: February 12, 2008 Authors: Joe Richardsi, David Hartmanii Document Version: 2.1

IBM System Storage DS3300 Storage System 1000 Mailbox Clustered Continuous Replication

Microsoft Exchange 2007 Storage Solution

Page 2 Joe Richards, David Hartman Copyright IBM Corp. 2008

Content IBM System Storage DS3300 Storage System 1000 Mailbox Clustered Continuous Replication Microsoft Exchange 2007 Storage Solution ............. 1Content ............................................................................................. 2Overview ........................................................................................... 3Disclaimer .......................................................................................... 3Features ............................................................................................ 3Solution Description ............................................................................ 4

Targeted Customer Profile ................................................................. 6Tested Deployment ........................................................................... 6

Simulated Exchange Configuration: .................................................. 6Primary Storage Hardware .............................................................. 7Primary Storage Software ............................................................... 7Primary Storage Disk Configuration (Mailbox Store Disks) ................... 8Primary Storage Disk Configuration (Transactional Log Disks) .............. 8

Replication Configuration ................................................................... 8Replicated Storage Hardware .......................................................... 9Replicated Storage Software ........................................................... 9Replicated Storage Disk Configuration (Mailbox Store Disks) ................ 9Replicated Storage Disk Configuration (Transactional Log Disks) ......... 10

Best Practices .................................................................................... 10Core Storage/Clustered Continuous Replication .................................... 10Backup strategy .............................................................................. 12

Contact for Additional Information ........................................................ 12Test Results Summary ........................................................................ 12

Reliability ....................................................................................... 12Primary Storage Performance Results ................................................. 13Streaming Backup Performance ......................................................... 13

Database Read-only Performance .................................................... 14Log Read-only Performance ............................................................ 14

Conclusion ........................................................................................ 14Appendix A Test Results ...................................................................... 15

Stress Test Result Report ................................................................. 15Performance Test Result Report ......................................................... 17Streaming backup Test Result Report ................................................. 19SoftRecovery Test Result Report ........................................................ 21




Overview This document provides information on the IBM System Storage DS3300 iSCSI storage solution for Microsoft Exchange Server2007, based upon the Microsoft Exchange Solution Reviewed Program (ESRP) Storage program*. *The ESRP Storage program was developed by Microsoft Corporation to provide a common storage testing framework for vendors to provide information on their storage solutions for Microsoft Exchange Server 2007 software. For more details on the Microsoft ESRP Storage program, please visit: http://www.microsoft.com/technet/prodtechnol/exchange/2007/esrp.mspx Disclaimer This document has been produced independently of Microsoft Corporation. Microsoft Corporation expressly disclaims responsibility for, and makes no warranty, express or implied, with respect to, the accuracy of the contents of this document. Please also note IBM disclaimer at end of this document.

Features This document describes an Exchange storage solution for 1000 users on the IBM System Storage DS3300 storage system using Microsofts Clustered Continuous Replication technology. The tested user profile was 0.384 IOPS (0.32 + 20% overhead) per user with a mailbox size of 500 MB. The solution can be scaled to support additional users and/or larger mailboxes by increasing the number of disks to support the increased I/O load & space requirements. The IBM System Storage DS3300 storage system is an excellent choice for small to medium sized businesses requiring scalable, highly-available, and high-performance iSCSI storage. As can be observed from the reports included in this paper, both the System Storage DS3300 and the System x 3950, as configured, could have easily handled a much larger Exchange Server 2007 workload. This could consist of larger mailbox sizes, additional users, or both. The IBM System Storage DS3300 is an ideal storage platform for Exchange Server 2007 data. With iSCSI host attach capabilities, system administrators can migrate from internal storage using familiar Ethernet technology & topology principals. Cost of deployment can be significantly lower than storage solutions based on other connectivity protocols such as fibre-channel, and SAS. Scalability and deployment constraints can also be eased when compared to direct-attach storage (DAS) solutions such as SAS.




The IBM System Storage DS3300 can be equipped with one or two controllers, each with up to two 1GB Ethernet host adapter ports. Each controller can be configured with up to 1GB cache, and the base unit can support up to 12 SAS or SATA disk drives. Expansion is achieved via the EXP3000, each base unit can connect to up to three EXP3000 drawers, for a maximum capacity of 48 TB using 1TB SATA drives. Configuration and management complexity of the storage array is reduced such that common tasks are handled by wizards, which can help administrators get the system up & running quickly. Full Microsoft VSS, VDS, and MPIO support further ease Exchange Server 2007 backup, recovery, and high-availability configuration. For more information on the IBM System Storage DS3300, please visit: http://www-03.ibm.com/systems/storage/disk/ds3000/ds3300/ The listing for the IBM System Storage DS3300 in the Windows Server Catalog can be found by clicking on the following link: http://www.windowsservercatalog.com/item.aspx?idItem=2c0ed642-fae8-c06b-2853-f812d7ae03b2 The IBM System x3950 M2 is highly-scalable enterprise-class server suited for the most demanding workloads. With up to 64 cores, and 1TB of RAM, the System x3950 is perfect for scale-up Exchange Server 2007 consolidations and migrations. Expandable from one to four nodes, this server is designed to grow with your business, which helps to maximize investments made in server infrastructure. For additional information on the IBM System x3950 M2, please visit: http://www-03.ibm.com/systems/x/hardware/enterprise/x3950m2/index.html

Solution Description This is a Clustered Continuous Replication (CCR) solution. The tested configuration is comprised of two identical of IBM System x servers. We used one IBM System Storage DS3300 storage system with a dual controller module attached to 2 EXP3000 storage expansion drive enclosures, for a total of 36 disks, and one IBM System x 3950 server (4 x 3.0 GHz processors, 16GB RAM) for each node of the 2-node cluster.




Each IBM System x3950 server was directly attached to the IBM System Storage DS3300 storage enclosure on the primary site. The secondary site was configured to be identical to the primary. The primary and secondary storage do not share storage array controllers or disks.

Figure 1: Basic CCR configuration (courtesy of Microsoft)

Cluster continuous replication (CCR) is a high availability feature of Microsoft Exchange Server 2007 that combines the asynchronous log shipping and replay technology built into Exchange 2007 with the failover and management features provided by the Cluster service. CCR is designed to provide high availability for Exchange 2007 Mailbox servers by providing a solution that:

Has no single point of failure. Has no special hardware requirements. Has no shared storage requirements. Can be deployed in one or two datacenter configurations. Can reduce full backup frequency, reduce total backed up data volume,

and shorten the service level agreement (SLA) for recovery time from first failure.

CCR uses the database failure recovery functionality in Exchange 2007 to enable the continuous and asynchronous updating of a second copy of a database with the changes that have been made to the active copy of the database. During installation of the passive node in a CCR environment, each storage group and its database is copied from the active node to the passive node. This operation is called seeding, and it provides a baseline of the database for replication. After the initial seeding is performed, log copying and replay are performed continuously.




The ESRP-Storage program focuses on storage solution testing to address performance and reliability issues with storage design. However, storage is not the only factor to take into consideration when designing a scale-up Exchange solution. Other factors which affect the server scalability are: server processor utilization, server physical and virtual memory limitations, resource requirements for other applications, directory and network service latencies, network infrastructure limitations, replication and recovery requirements, and client usage profiles. All these factors are beyond the scope for ESRP-Storage. Therefore, the number of mailboxes hosted per server as part of the tested configuration may not necessarily be viable for some customer deployments.

For more information on identifying and addressing performance bottlenecks in an Exchange Server 2007 system, please refer to Microsoft's Troubleshooting Microsoft Exchange Server Performance document, available at:

http://go.microsoft.com/fwlink/?LinkId=23454

Targeted Customer Profile This solution is intended for small organizations hosting up to 1000 Exchange Server 2007 mailboxes. The configuration used for testing follows:

Number of mailboxes: 1000 Number of hosts attached to the storage system: 1 (per cluster node) User I/O profile: 0.32 I/O Operation per second with 20% headroom

(Tested 0.384 IOPS) 1 Storage Group, 3 Databases (per cluster node) Mailbox Size: 500 MB Backup strategy: N/A Clustered Continuous Replication (CCR) implemented (Exchange

Server 2007 controlled)

Tested Deployment The following tables summarize the testing environment:

Simulated Exchange Configuration: Number of Exchange mailboxes simulated

1000

Number of hosts 1 (per cluster node) Number of storage groups/host 3 Number of mailbox stores/storage group

1

Number of mailboxes/mailbox store 333




Number of mailbox store LUNs/storage group

3

Simulated profile: I/Os per second per mailbox (IOPS, include 20% headroom)

0.384 IOPS (.32 +20%)

Database LUN size 400 GB (x3) Log LUN size 136 GB (x3) Backup LUN size/storage group N/A Total database size for performance testing

488.4 GB

% storage capacity used by Exchange database**

41%

**Storage performance characteristics change based on the percentage utilization of the individual disks. Tests that use a small percentage of the storage (~25%) may exhibit reduced throughput if the storage capacity utilization is significantly increased beyond what is tested in this paper.

Primary Storage Hardware Storage connectivity (Fiber Channel, SAS, SATA, iSCSI)

iSCSI

Storage model and OS/firmware revision

DS3300 Firmware Version 96.70.03.00

Storage cache 4GB of battery-backed, mirrored Number of storage controllers 2 Number of storage ports 2 Maximum bandwidth of storage connectivity to host

6 GB (2x3GB HBA ports)

HBA model and firmware QLogic 4032 Firmware Ver 3.1.3.25 (Jumbo Frames, iSCSI offload and TOE offload not used or configured)

Number of HBAs/host 1 (Dual Port) Total number of disks tested in solution

36 (3 5x5 Raid 10 (DB), 3 Raid 1 (Logs))

Maximum number of spindles can be hosted in the storage

48 (with 3 Expansion Enclosures)

Primary Storage Software HBA driver QLogic 2.1.4.18 HBA Queue Target Setting N/A HBA Queue Depth Setting 32 Multi-Pathing Engenio Information Technologies

MPIO Device Specific Module




v01.01.35.52, Fail-over driver Host OS Microsoft Windows Server 2003 R2

Enterprise x64 Edition w/ Service Pack 2

ESE.dll file version 8.0.685.24 Replication solution name/version CCR

Primary Storage Disk Configuration (Mailbox Store Disks) Disk type, speed and firmware revision

SCSI SAS 136.7 GB, 10K, FW-MS02

Raw capacity per disk (GB) 136.7 Number of physical disks in test 30 total raw storage capacity (GB) 4101 Number of slices per LUN or number of disks per LUN

N/A

Raid level Raid 10, storage level Total formatted capacity 1197 GB Storage capacity utilization 29.2% Database capacity utilization 23.8%

Primary Storage Disk Configuration (Transactional Log Disks) Disk type, speed and firmware revision


Raw capacity per disk (GB) 136.7 Number of spindles in test 6 Total raw storage capacity (GB) 820.2 Number of slices per LUN or number of disks per LUN

N/A

Raid level Raid 1, storage level Total formatted capacity 410.1

Replication Configuration Replication mechanism Cluster Continuous Replication Number of links 2 Simulated link distance LAN Link type IP Link bandwidth GigE (1 Gbps)




Replicated Storage Hardware Storage connectivity (Fiber Channel, SAS, SATA, iSCSI)

iSCSI

Storage model and OS/firmware revision

DS3300 Firmware Version 96.70.03.00

Storage cache 4GB of battery-backed, mirrored Number of storage controllers 2 Number of storage ports 2 Maximum bandwidth of storage connectivity to host

6 GB (2x3GB HBA ports)

HBA model and firmware QLogic 4032 Firmware Version 3.1.3.25 (Jumbo Frames, iSCSI offload and TOE offload not used or configured)

Number of HBAs/host 1 (Dual port) Total number of disks tested in solution

36

Maximum number of spindles can be hosted in the storage

48

Replicated Storage Software HBA driver QLogic 2.1.4.18 HBA Queue Target Setting N/A HBA Queue Depth Setting 32 Multi-Pathing Engenio Information Technologies

MPIO Device Specific Module v01.01.35.52, Fail-over driver

Host OS Microsoft Windows Server 2003 R2 Enterprise x64 Edition w/ Service Pack 2

ESE.dll file version 8.0.685.24 Replication solution name/version CCR

Replicated Storage Disk Configuration (Mailbox Store Disks) Disk type, speed and firmware revision

SAS 136.7 GB, 10K, FW-MS02

Raw capacity per disk (GB) 136.7 Number of physical disks in test 30 total raw storage capacity (GB) 4101 Number of slices per LUN or number of disks per LUN

N/A

Raid level Raid 10, storage level Total formatted capacity 1197 GB




Storage capacity utilization 29.2% Database capacity utilization 23.8%

Replicated Storage Disk Configuration (Transactional Log Disks) Disk type, speed and firmware revision


Raw capacity per disk (GB) 136.7 Number of Spindles in test 6 Total raw storage capacity (GB) 820.2 Number of slices per LUN or number of disks per LUN

N/A

Raid level Raid 1, storage level Total formatted capacity 410.1

Best Practices Exchange Server 2007 is a disk-intensive application. Based on the testing run using the ESRP Storage framework, we would recommend the following to improve the storage performance. For Exchange Server 2007 best practices on storage design, please visit: http://technet.microsoft.com/en-us/library/bb124518.aspx

Core Storage/Clustered Continuous Replication 1. Use the Microsoft Diskpart utility to align the sectors of all Exchange Server 2007 storage volumes prior to formatting. For detailed steps about how to use Diskpart to align I/O with storage track boundaries, visit: http://technet.microsoft.com/en-us/library/aa998219.aspx 2. Configure the NTFS volumes hosting your Exchange Server 2007 databases with an NTFS allocation unit size of 64KB. The recommendation of 64KB is based on the performance improvements seen with large sequential read operations. 3. Isolate Exchange Server 2007 volumes from other disk intensive applications. For the purpose of this testing, the storage subsystem was dedicated exclusively to Exchange Server 2007. Sharing the storage with other applications may have a negative impact on Exchange Server 2007 performance.




4. Isolate Exchange Server 2007 database and log disk I/O on separate physical volume groups. Separate the transaction log files (sequential I/O) from databases (random I/O) to maximize I/O performance and increase fault tolerance. Storing sequentially accessed files separately keeps the disk heads in position for sequential I/O, which reduces the amount of time required to locate data. From a recoverability perspective, separating a storage group's transaction logs and databases ensure that a catastrophic failure of a particular set of physical disks will not cause a loss of both database and transaction logs. 5. The cache block size is a global parameter for the storage subsystem. Set the cache block size closest to the typical I/O size, which is 16K size for large block and sequential I/O. 6. Configure the entire capacity of a volume group in a single volume (LUN). Multiple LUNs on one volume group typically increases the seek time penalty. 7. Use all available host-side channels. Switch zoning and multi-pathing HBA drivers are all useful tools to ensure that all channels are kept active. 8. Balance I/O across the dual controllers of the storage system and strive to keep both controllers busy. Locate database volume and their corresponding log volume on separate paths. 9. Create volume groups across drive trays to distribute I/O across drive-side loops. 10. Choose faster disks. A 15K RPM drive yields ~15% more performance than a 10K RPM drive. 11. For the log LUNs, use RAID 10 volumes and enable write caching with mirroring. 12. For the database LUNS, enable read caching and write caching with mirroring. 13. Troubleshoot any latency issues on the DS3300 storage system using standard tools such as Windows Performance Monitor, or the IBM Storage Manager Performance Monitor. These tools can help determine where the heaviest disk I/O reads and writes are occurring, and what the transfer rates are. Most Exchange storage latency issues are caused by an insufficient number of disks for a given I/O load. The number of IOPS hitting a given drive can be measured using the Performance Monitor Logical Disk object, Disk Transfers/sec counter.




14. Ensure that baselines are taken on all attached servers and arrays. Setup a private LAN for the arrays and servers. Do not use your public LAN. If you must use your public LAN, build a VLAN to handle storage-specific traffic. Look into utilizing maximum throughput with Jumbo Frames. 15. Because of the GigE iSCSI communication between the server(s) and Array(s), basic network troubleshooting should find most issues. Ensure network settings are the same for all devices. Some settings to take into account are:

a. Speed and duplex (Do not use Auto) b. Jumbo frames c. Spanning-Tree Protocol d. Flow control e. Unicast Storm control f. Ethernet priority g. CHAP (Without CHAP, any host can access your storage array(s))

16. See the following Microsoft documentation for storage based replication best practices and support criteria:

Planning for Cluster Continuous Replication

Multi-site data replication support for Exchange

Backup strategy N/A

Contact for Additional Information https://www-912.ibm.com/ssg/tsfeedback.nsf/feedback?OpenForm

Test Results Summary This section provides a high level summary of the test data from ESRP Storage and the links to the detailed html reports which are generated by ESRP Storage testing framework. Please click on the underlined headings below to view the html report for each test.

Reliability A number of tests in the ESRP Storage framework are designed to check storage subsystem reliability. Reliability tests runs for 24 hours. The goal is to verify the storage subsystem can handle high Exchange Server 2007 I/O loads for an extended period of time. Both log and database files are analyzed for integrity after the stress test to ensure there is no database and/or log corruption.




The following list provides an overview: (clicking on the underlined word will show the relevant html report generated after the reliability tests ran)

No errors reported in the saved eventlog file. No errors reported in during the database and log checksum processes.

Primary Storage Performance Results The Primary Storage performance testing is designed to exercise the storage with the maximum sustainable Exchange type of I/O for 2 hours. The test is designed to show how long it takes for the storage subsystem to respond to Exchange Server 2007 I/O under load. The data below is the sum of all of the logical disk I/Os and average of all the logical disk I/O latencies during the 2 hours test duration. Each server is listed separately and the aggregate numbers across all servers are listed as well. Individual Server Metrics:

The sum of I/Os across Storage Groups and the average latency across all Storage Groups on a per server basis. Database I/O Average Database Disk Transfers/sec 1858.629 Average Database Disk Reads/sec 1027.195 Average Database Disk Writes/sec 831.434 Average Database Disk Read Latency (ms)

0.017

Average Database Disk Write Latency (ms)

0.014

Transaction Log I/O Average Log Disk Writes/sec 82.791 Average Log Disk Write Latency (ms) 6.6 ms

Streaming Backup Performance For the Version 1.0 release, only the streaming backup type is supported for testing within the ESRP Storage framework. There are two tests in this section. The first one is designed to measure the read I/O performance metrics by running checksums on all the databases and log files. The second test is designed to measure the end-to-end performance when the databases are backed up to disks.




Database Read-only Performance The test is to measure the maximum rate at which databases were backed-up via streaming method. The following table shows the average rate for a single database file. MB read/sec per storage group 40.213 MB read/sec total 120.639

Log Read-only Performance The test is designed to measure the maximum rate at which the log files were replayed against the databases. The following table shows the average rate for 500 log files replayed in a single storage group. Each log file is 1 MB in size. Average time to play one log file (sec)

0.434

Average log disks Read Bytes/sec 75.1

Conclusion This document is developed by IBM, and reviewed by Microsoft Exchange Product team. The test results/data presented in this document are based upon the tests introduced in the ESRP Storage test framework. Customers should not use the data solely for pre-deployment verification. It is still necessary to go through the requisite exercises to validate the storage subsystem design for each specific customer environment. The ESRP Storage program is not designed to be a benchmarking program; tests are not designed toward obtaining the maximum throughput for a given solution. Rather, the ESRP Storage program is focused on producing recommendations from storage subsystem vendors like IBM for the Exchange Server 2007 application. So the data presented in this document should not be used for direct comparisons among the various vendor solutions.




Appendix A Test Results

Stress Test Result Report Test Summary

Overall Test Result Pass Machine Name TUMBLEWEED Test Description Test Start Time 2/11/2008 6:05:03 PM Test End Time 2/12/2008 6:07:36 PM Jetstress Version 08.01.0177.000 Ese Version 08.00.0685.024 Operating System Microsoft Windows Server 2003 R2 Service Pack 2

(5.2.3790.131072) Performance Log C:\Program Files\Exchange

Jetstress\stress\Stress_2008_2_11_18_5_11.blg C:\Program Files\Exchange Jetstress\stress\DBChecksum_2008_2_12_18_7_36.blg

Database Sizing and Throughput

Achieved I/O per Second 1779.048 Target I/O per Second 384 Initial database size 541618651136 Final database size 717791084544 Database files (count) 3 Jetstress System Parameters

Thread count 32 (per-storage group) Log buffers 9000 Minimum database cache 96.0 MB Maximum database cache 768.0 MB Insert operations 25% Delete operations 10% Replace operations 50% Read operations 15% Lazy commits 80%




Disk Subsystem Performance

LogicalDisk Avg. Disk sec/Read

Avg. Disk sec/Write

Disk Reads/sec Disk Writes/sec

Avg. Disk Bytes/Write

Database (F:) 0.015 0.013 335.459 260.227 (n/a) Database (G:) 0.019 0.014 336.421 255.856 (n/a) Database (H:) 0.019 0.014 335.230 255.855 (n/a) Log (I:) 0.000 0.008 0.000 73.013 24833.316 Log (J:) 0.000 0.006 0.000 78.500 23190.660 Log (K:) 0.000 0.006 0.000 78.580 23232.247 Host System Performance Counter Average Minimum Maximum % Processor Time 2.290 0.000 4.781 Available MBytes 14453.098 14431.000 15174.000 Free System Page Table Entries 16741734.913 16741046.000 16741761.000 Transition Pages RePurposed/sec 0.000 0.000 0.000 Pool Nonpaged Bytes 68578473.956 68423680.000 68956160.000 Pool Paged Bytes 51407036.444 49901568.000 54140928.000 Database Page Fault Stalls/sec 0.000 0.000 0.333 Test Log 2/11/2008 6:05:03 PM -- Jetstress testing begins ... 2/11/2008 6:05:03 PM -- Prepare testing begins ... 2/11/2008 6:05:07 PM -- Attaching databases ... 2/11/2008 6:05:07 PM -- Prepare testing ends. 2/11/2008 6:05:07 PM -- Dispatching transactions begins ... 2/11/2008 6:05:07 PM -- Database cache settings: (minimum: 96.0 MB, maximum: 768.0 MB) 2/11/2008 6:05:07 PM -- Database flush thresholds: (start: 7.7 MB, stop: 15.4 MB) 2/11/2008 6:05:11 PM -- Database read latency thresholds: (average: 0.02 seconds/read, maximum: 0.1 seconds/read). 2/11/2008 6:05:11 PM -- Log write latency thresholds: (average: 0.01 seconds/write, maximum: 0.1 seconds/write). 2/11/2008 6:05:12 PM -- Operation mix: Sessions 32, Inserts 25%, Deletes 10%, Replaces 50%, Reads 15%, Lazy Commits 80%. 2/11/2008 6:05:12 PM -- Performance logging begins (interval: 15000 ms). 2/11/2008 6:05:12 PM -- Attaining prerequisites: 2/11/2008 6:07:25 PM -- \Database(JetstressWin)\Database Cache Size, Last: 728334300.0 (lower bound: 724775700.0, upper bound: none) 2/12/2008 6:07:26 PM -- Performance logging ends. 2/12/2008 6:07:26 PM -- JetInterop batch transaction stats: 1398742, 1399412, and 1398233. 2/12/2008 6:07:27 PM -- Dispatching transactions ends. 2/12/2008 6:07:27 PM -- Shutting down databases ... 2/12/2008 6:07:36 PM -- Instance1988.1 (complete), Instance1988.2 (complete), and Instance1988.3 (complete) 2/12/2008 6:07:37 PM -- Performance logging begins (interval: 30000 ms). 2/12/2008 6:07:37 PM -- Verifying database checksums ... 2/12/2008 7:18:52 PM -- F: (100% processed), G: (100% processed), and H: (100% processed) 2/12/2008 7:18:53 PM -- Performance logging ends. 2/12/2008 7:18:53 PM -- C:\Program Files\Exchange Jetstress\stress\DBChecksum_2008_2_12_18_7_36.blg has 142 samples. 2/12/2008 7:19:07 PM -- C:\Program Files\Exchange




Jetstress\stress\DBChecksum_2008_2_12_18_7_36.html is saved. 2/12/2008 7:19:07 PM -- Verifying log checksums ... 2/12/2008 7:19:16 PM -- I:\ (37 logs passed), J:\ (32 logs passed), and K:\ (34 logs passed) 2/12/2008 7:19:16 PM -- C:\Program Files\Exchange Jetstress\stress\Stress_2008_2_11_18_5_11.blg has 5768 samples. 2/12/2008 7:19:16 PM -- Creating test report ... 2/12/2008 7:23:15 PM -- Volume F: has 0.0150 for Avg. Disk sec/Read. 2/12/2008 7:23:15 PM -- Volume G: has 0.0188 for Avg. Disk sec/Read. 2/12/2008 7:23:15 PM -- Volume H: has 0.0188 for Avg. Disk sec/Read. 2/12/2008 7:23:15 PM -- Volume I: has 0.0080 for Avg. Disk sec/Write. 2/12/2008 7:23:15 PM -- Volume I: has 0.0000 for Avg. Disk sec/Read. 2/12/2008 7:23:15 PM -- Volume J: has 0.0062 for Avg. Disk sec/Write. 2/12/2008 7:23:15 PM -- Volume J: has 0.0000 for Avg. Disk sec/Read. 2/12/2008 7:23:15 PM -- Volume K: has 0.0062 for Avg. Disk sec/Write. 2/12/2008 7:23:15 PM -- Volume K: has 0.0000 for Avg. Disk sec/Read. 2/12/2008 7:23:15 PM -- Test has 0.333334382507845 Maximum Database Page Fault Stalls/sec. 2/12/2008 7:23:15 PM -- Test has 0 Database Page Fault Stalls/sec samples higher than 0. 2/12/2008 7:23:15 PM -- C:\Program Files\Exchange Jetstress\stress\Stress_2008_2_11_18_5_11.xml has 5759 samples queried.

Performance Test Result Report Test Summary Overall Test Result

Pass

Machine Name TUMBLEWEED Test Description Test Start Time 2/11/2008 2:41:33 PM Test End Time 2/11/2008 4:44:16 PM Jetstress Version 08.01.0177.000 Ese Version 08.00.0685.024 Operating System Microsoft Windows Server 2003 R2 Service Pack 2

(5.2.3790.131072) Performance Log C:\Program Files\Exchange

Jetstress\performance\Performance_2008_2_11_14_41_41.blg C:\Program Files\Exchange Jetstress\performance\DBChecksum_2008_2_11_16_44_16.blg

Database Sizing and Throughput

Achieved I/O per Second 1858.629 Target I/O per Second 384 Initial database size 524296175616 Final database size 541618651136 Database files (count) 3 Jetstress System Parameters

Thread count 32 (per-storage group) Log buffers 9000 Minimum database cache 96.0 MB




Maximum database cache 768.0 MB Insert operations 25% Delete operations 10% Replace operations 50% Read operations 15% Lazy commits 80% Disk Subsystem Performance


Avg. Disk sec/Write

Disk Reads/sec

Disk Writes/sec


Database (F:)

0.015 0.014 340.738 279.544 (n/a)

Database (G:)

0.018 0.014 342.661 276.232 (n/a)

Database (H:)

0.018 0.014 343.796 275.658 (n/a)

Log (I:) 0.000 0.008 0.001 78.566 26012.950 Log (J:) 0.000 0.006 0.000 84.847 24355.998 Log (K:) 0.000 0.006 0.001 84.959 24196.667 Host System Performance

Counter Average Minimum Maximum % Processor Time 2.627 0.977 3.841 Available MBytes 14320.031 14239.000 14331.000 Free System Page Table Entries 16741730.048 16741249.000 16741761.000 Transition Pages RePurposed/sec 0.000 0.000 0.000 Pool Nonpaged Bytes 68388420.267 68362240.000 68440064.000 Pool Paged Bytes 53405081.600 52785152.000 54394880.000 Database Page Fault Stalls/sec 0.000 0.000 0.000 Test Log 2/11/2008 11:37:08 AM -- Prepare testing begins ... 2/11/2008 11:37:08 AM -- Creating F:\Jetstress1.edb. 2/11/2008 11:37:08 AM -- Database cache settings: (minimum: 32.0 MB, maximum: 256.0 MB) 2/11/2008 11:37:08 AM -- Database flush thresholds: (start: 2.6 MB, stop: 5.1 MB) 2/11/2008 12:30:31 PM -- 60.0% of 162.8 GB complete (5951010 records inserted). 2/11/2008 1:13:55 PM -- 100.0% of 162.8 GB complete (9973241 records inserted). 2/11/2008 1:13:58 PM -- Duplicating 2 databases: 2/11/2008 2:16:35 PM -- 100.0% of 325.5 GB complete (325.5 GB duplicated). 2/11/2008 2:16:38 PM -- Attaching databases ... 2/11/2008 2:16:38 PM -- Prepare testing ends. 2/11/2008 2:41:33 PM -- Jetstress testing begins ... 2/11/2008 2:41:33 PM -- Prepare testing begins ... 2/11/2008 2:41:37 PM -- Attaching databases ...




2/11/2008 2:41:37 PM -- Prepare testing ends. 2/11/2008 2:41:37 PM -- Dispatching transactions begins ... 2/11/2008 2:41:37 PM -- Database cache settings: (minimum: 96.0 MB, maximum: 768.0 MB) 2/11/2008 2:41:37 PM -- Database flush thresholds: (start: 7.7 MB, stop: 15.4 MB) 2/11/2008 2:41:41 PM -- Database read latency thresholds: (average: 0.02 seconds/read, maximum: 0.05 seconds/read). 2/11/2008 2:41:41 PM -- Log write latency thresholds: (average: 0.01 seconds/write, maximum: 0.05 seconds/write). 2/11/2008 2:41:42 PM -- Operation mix: Sessions 32, Inserts 25%, Deletes 10%, Replaces 50%, Reads 15%, Lazy Commits 80%. 2/11/2008 2:41:42 PM -- Performance logging begins (interval: 15000 ms). 2/11/2008 2:41:42 PM -- Attaining prerequisites: 2/11/2008 2:44:07 PM -- \Database(JetstressWin)\Database Cache Size, Last: 728399900.0 (lower bound: 724775700.0, upper bound: none) 2/11/2008 4:44:08 PM -- Performance logging ends. 2/11/2008 4:44:08 PM -- JetInterop batch transaction stats: 133274, 133645, and 133928. 2/11/2008 4:44:09 PM -- Dispatching transactions ends. 2/11/2008 4:44:09 PM -- Shutting down databases ... 2/11/2008 4:44:16 PM -- Instance3832.1 (complete), Instance3832.2 (complete), and Instance3832.3 (complete) 2/11/2008 4:44:17 PM -- Performance logging begins (interval: 15000 ms). 2/11/2008 4:44:17 PM -- Verifying database checksums ... 2/11/2008 5:38:01 PM -- F: (100% processed), G: (100% processed), and H: (100% processed) 2/11/2008 5:38:02 PM -- Performance logging ends. 2/11/2008 5:38:02 PM -- C:\Program Files\Exchange Jetstress\performance\DBChecksum_2008_2_11_16_44_16.blg has 214 samples. 2/11/2008 5:38:22 PM -- C:\Program Files\Exchange Jetstress\performance\DBChecksum_2008_2_11_16_44_16.html is saved. 2/11/2008 5:38:22 PM -- Verifying log checksums ... 2/11/2008 5:38:31 PM -- I:\ (35 logs passed), J:\ (34 logs passed), and K:\ (37 logs passed) 2/11/2008 5:38:31 PM -- C:\Program Files\Exchange Jetstress\performance\Performance_2008_2_11_14_41_41.blg has 489 samples. 2/11/2008 5:38:31 PM -- Creating test report ... 2/11/2008 5:38:52 PM -- Volume F: has 0.0145 for Avg. Disk sec/Read. 2/11/2008 5:38:52 PM -- Volume G: has 0.0176 for Avg. Disk sec/Read. 2/11/2008 5:38:52 PM -- Volume H: has 0.0177 for Avg. Disk sec/Read. 2/11/2008 5:38:52 PM -- Volume I: has 0.0077 for Avg. Disk sec/Write. 2/11/2008 5:38:52 PM -- Volume I: has 0.0000 for Avg. Disk sec/Read. 2/11/2008 5:38:52 PM -- Volume J: has 0.0061 for Avg. Disk sec/Write. 2/11/2008 5:38:52 PM -- Volume J: has 0.0000 for Avg. Disk sec/Read. 2/11/2008 5:38:52 PM -- Volume K: has 0.0061 for Avg. Disk sec/Write. 2/11/2008 5:38:52 PM -- Volume K: has 0.0001 for Avg. Disk sec/Read. 2/11/2008 5:38:52 PM -- Test has 0 Maximum Database Page Fault Stalls/sec. 2/11/2008 5:38:52 PM -- Test has 0 Database Page Fault Stalls/sec samples higher than 0. 2/11/2008 5:38:52 PM -- C:\Program Files\Exchange Jetstress\performance\Performance_2008_2_11_14_41_41.xml has 479 samples queried.

Streaming backup Test Result Report Streaming Backup Statistics - All

Database Instance

Database Size (MBytes)

Elapsed Backup Time

MBytes Transferred/sec

Instance3532.1 166667.27 00:52:28 52.93




Instance3532.2 166667.27 01:22:02 33.86 Instance3532.3 166667.27 01:22:03 33.85 Jetstress System Parameters

Thread count 32 (per-storage group) Log buffers 9000 Minimum database cache 96.0 MB Maximum database cache 768.0 MB Insert operations 25% Delete operations 10% Replace operations 50% Read operations 15% Lazy commits 80% Disk Subsystem Performance


Avg. Disk sec/Write

Disk Reads/sec

Disk Writes/sec


Database (f:)

0.001 0.000 270.546 0.012 (n/a)

Database (g:)

0.004 0.000 270.851 0.003 (n/a)

Database (h:)

0.004 0.000 270.799 0.003 (n/a)

Log (i:) 0.000 0.000 0.000 0.010 142.341 Log (j:) 0.000 0.000 0.000 0.001 10.334 Log (k:) 0.000 0.000 0.000 0.001 14.539 Host System Performance

Counter Average Minimum Maximum % Processor Time 2.102 1.452 2.669 Available MBytes 15237.582 15080.000 15259.000 Free System Page Table Entries 16741762.341 16741704.000 16741768.000 Transition Pages RePurposed/sec 0.000 0.000 0.000 Pool Nonpaged Bytes 68269118.439 68255744.000 68313088.000 Pool Paged Bytes 54145448.585 52531200.000 55554048.000 Database Page Fault Stalls/sec 0.000 0.000 0.000




Test Log2/13/2008 10:55:11 AM -- Prepare testing begins ... 2/13/2008 10:55:39 AM -- Prepare testing begins ... 2/13/2008 10:55:39 AM -- Creating f:\Jetstress1.edb. 2/13/2008 10:55:39 AM -- Database cache settings: (minimum: 32.0 MB, maximum: 256.0 MB) 2/13/2008 10:55:39 AM -- Database flush thresholds: (start: 2.6 MB, stop: 5.1 MB) 2/13/2008 11:48:13 AM -- 60.0% of 162.8 GB complete (5950891 records inserted). 2/13/2008 12:32:10 PM -- 100.0% of 162.8 GB complete (9973230 records inserted). 2/13/2008 12:32:13 PM -- Duplicating 3 databases: 2/13/2008 2:19:23 PM -- 100.0% of 488.3 GB complete (488.3 GB duplicated). 2/13/2008 2:19:27 PM -- Attaching databases ... 2/13/2008 2:19:27 PM -- Prepare testing ends. 2/13/2008 2:39:32 PM -- Jetstress testing begins ... 2/13/2008 2:39:32 PM -- Prepare testing begins ... 2/13/2008 2:39:37 PM -- Attaching databases ... 2/13/2008 2:39:37 PM -- Prepare testing ends. 2/13/2008 2:39:41 PM -- Performance logging begins (interval: 15000 ms). 2/13/2008 2:39:41 PM -- Streaming backup databases ... 2/13/2008 4:01:46 PM -- Performance logging ends. 2/13/2008 4:01:46 PM -- Instance3532.1 (100% processed), Instance3532.2 (100% processed), and Instance3532.3 (100% processed) 2/13/2008 4:01:46 PM -- C:\Program Files\Exchange Jetstress\streamingbackup\StreamingBackup_2008_2_13_14_39_37.blg has 328 samples. 2/13/2008 4:01:46 PM -- Creating test report ...

SoftRecovery Test Result Report Soft-Recovery Statistics - All Database Instance Log files replayed Elapsed seconds Instance2456.1 501 209.515625 Instance2456.2 508 225.515625 Instance2456.3 514 225.265625




Disk Subsystem Performance


Avg. Disk sec/Write

Disk Reads/sec

Disk Writes/sec


Database (f:)

0.054 0.028 918.577 9.304 (n/a)

Database (g:)

0.054 0.026 889.231 9.147 (n/a)

Database (h:)

0.055 0.022 855.429 9.254 (n/a)

Log (i:) 0.003 0.003 74.457 2.733 3544.956 Log (j:) 0.003 0.002 74.905 2.400 3513.713 Log (k:) 0.003 0.002 75.938 2.427 4715.628 Host System Performance

Counter Average Minimum Maximum % Processor Time 2.660 1.136 4.395 Available MBytes 14499.813 14383.000 15180.000 Free System Page Table Entries 16741759.875 16741635.000 16741761.000 Transition Pages RePurposed/sec 0.000 0.000 0.000 Pool Nonpaged Bytes 68430482.286 68345856.000 68612096.000 Pool Paged Bytes 52922002.286 52641792.000 53170176.000 Database Page Fault Stalls/sec 0.005 0.000 0.500 Test Log2/12/2008 7:45:42 PM -- Jetstress testing begins ... 2/12/2008 7:45:42 PM -- Prepare testing begins ... 2/12/2008 7:45:45 PM -- Attaching databases ... 2/12/2008 7:45:45 PM -- Prepare testing ends. 2/12/2008 7:45:45 PM -- Dispatching transactions begins ... 2/12/2008 7:45:45 PM -- Database cache settings: (minimum: 96.0 MB, maximum: 768.0 MB) 2/12/2008 7:45:45 PM -- Database flush thresholds: (start: 7.7 MB, stop: 15.4 MB) 2/12/2008 7:45:49 PM -- Database read latency thresholds: (average: 0.02 seconds/read, maximum: 0.05 seconds/read). 2/12/2008 7:45:49 PM -- Log write latency thresholds: (average: 0.01 seconds/write, maximum: 0.05 seconds/write). 2/12/2008 7:45:50 PM -- Operation mix: Sessions 32, Inserts 25%, Deletes 10%, Replaces 50%, Reads 15%, Lazy Commits 80%. 2/12/2008 7:45:50 PM -- Performance logging begins (interval: 15000 ms). 2/12/2008 7:45:50 PM -- Generating log files ... 2/12/2008 7:57:06 PM -- i:\ (100.0% generated), j:\ (100.0% generated), and k:\ (100.0% generated) 2/12/2008 7:57:07 PM -- Performance logging ends. 2/12/2008 7:57:07 PM -- JetInterop batch transaction stats: 9667, 9688, and 9815. 2/12/2008 7:57:08 PM -- Dispatching transactions ends. 2/12/2008 7:57:08 PM -- Shutting down databases ... 2/12/2008 7:57:15 PM -- Instance2456.1 (complete), Instance2456.2 (complete), and Instance2456.3 (complete) 2/12/2008 7:57:15 PM -- C:\Program Files\Exchange




Jetstress\softrecovery\Performance_2008_2_12_19_45_49.blg has 45 samples. 2/12/2008 7:57:15 PM -- Creating test report ... 2/12/2008 7:57:16 PM -- Volume f: has 0.0152 for Avg. Disk sec/Read. 2/12/2008 7:57:16 PM -- Volume g: has 0.0210 for Avg. Disk sec/Read. 2/12/2008 7:57:16 PM -- Volume h: has 0.0211 for Avg. Disk sec/Read. 2/12/2008 7:57:16 PM -- Volume i: has 0.0086 for Avg. Disk sec/Write. 2/12/2008 7:57:16 PM -- Volume i: has 0.0064 for Avg. Disk sec/Read. 2/12/2008 7:57:16 PM -- Volume j: has 0.0062 for Avg. Disk sec/Write. 2/12/2008 7:57:16 PM -- Volume j: has 0.0050 for Avg. Disk sec/Read. 2/12/2008 7:57:16 PM -- Volume k: has 0.0062 for Avg. Disk sec/Write. 2/12/2008 7:57:16 PM -- Volume k: has 0.0041 for Avg. Disk sec/Read. 2/12/2008 7:57:16 PM -- Test has 3.18961035026462 Maximum Database Page Fault Stalls/sec. 2/12/2008 7:57:16 PM -- Test has 0 Database Page Fault Stalls/sec samples higher than 0. 2/12/2008 7:57:16 PM -- C:\Program Files\Exchange Jetstress\softrecovery\Performance_2008_2_12_19_45_49.xml has 44 samples queried. 2/12/2008 7:57:19 PM -- C:\Program Files\Exchange Jetstress\softrecovery\Performance_2008_2_12_19_45_49.html is saved. 2/12/2008 7:57:20 PM -- Performance logging begins (interval: 2000 ms). 2/12/2008 7:57:20 PM -- Recovering databases ... 2/12/2008 8:01:07 PM -- Performance logging ends. 2/12/2008 8:01:07 PM -- Instance2456.1 (209.515625), Instance2456.2 (225.515625), and Instance2456.3 (225.265625) 2/12/2008 8:01:07 PM -- C:\Program Files\Exchange Jetstress\softrecovery\SoftRecovery_2008_2_12_19_57_19.blg has 112 samples. 2/12/2008 8:01:07 PM -- Creating test report ...




Copyright 2008 by International Business Machines Corporation. This document could include technical inaccuracies or typographical errors. IBM may make changes, improvements or alterations to the products, programs and services described in this document, including termination of such products, programs and services, at any time and without notice. Any statements regarding IBMs future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only. The information contained in this document is current as of the initial date of publication only, and IBM shall have no responsibility to update such information. Performance data for IBM and non-IBM products and services contained in this document was derived under specific operating and environmental conditions. The actual results obtained by any party implementing such products or services will depend on a large number of factors specific to such partys operating environment and may vary significantly. IBM makes no representation that these results can be expected or obtained in any implementation of any such products or services. THE INFORMATION IN THIS DOCUMENT IS PROVIDED AS-IS WITHOUT ANY WARRANTY, EITHER EXPRESS OR IMPLIED. IBM EXPRESSLY DISCLAIMS ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR INFRINGEMENT. References in this document to IBM products, programs, or services does not imply that IBM intends to make such products, programs or services available in all countries in which IBM operates or does business. Any reference to an IBM program or product in this document is not intended to state or imply that only that program or product may be used. Any functionally equivalent program or product, that does not infringe IBMs intellectually property rights, may be used instead. It is the users responsibility to evaluate and verify the operation of any non-IBM product, program or service. The provision of the information contained herein is not intended to, and does not grant any right or license under any IBM patents or copyrights. Inquiries regarding patent or copyright licenses should be made, in writing, to: IBM Director of Licensing IBM Corporation North Castle Drive Armonk, NY 10504-1785 U.S.A. IBM, the IBM logo, System x, and System Storage are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries or both. Microsoft and Windows are trademarks of Microsoft Corporation in the United States, other countries or both. i LSI Corporation, Beaverton, OR, [email protected] ii IBM Systems & Technology Group, Raleigh, NC, [email protected]

Date post:	24-Nov-2015
Category:	Documents
Upload:	subhendu-dhar
View:	25 times
Download:	0 times

IBM DS3300 ESRP Storage Solution Guide

Documents