IBM Lotus Notes Traveler 8.5.1 performance

This article reports IBM® Lotus® Notes® Traveler 8.5.1 performance results for Microsoft® Windows® 64-bit operating systems. In addition, it compares the performances of Lotus Notes Traveler 8.5 (64-bit) server and IBM Lotus Notes Traveler 8.5.1 (64-bit) server. Lotus Notes Traveler 8.5.1 for IBM Lotus Domino® 32-bit servers was not tested because large deployments of Lotus Notes Traveler server are preferred on Lotus Domino 64-bit servers.

Yuriy Veytsman, Staff Software Engineer, IBM

Yuriy Veytsman is a Staff Software Engineer on the IBM Lotus Domino server performance team. You can reach him at yuriy_veytsman@us.ibm.com.



Nirmala Venkatraman, Performance Architect, IBM

Nirmala Venkatraman is a Performance Architect on the Lotus Domino server performance team. You can reach her at nvenkatr@us.ibm.com.



Razeyah Stephen (razeyah@us.ibm.com ), Manager, Lotus Domino Performance Team, IBM

Razeyah Stephen is a manager of the IBM Lotus Domino and IBM Lotus Sametime performance teams. She joined IBM in October 1998.



09 March 2010

Also available in Chinese Russian Portuguese

Editor's note: Know a lot about this topic? Want to share your expertise? Participate in the IBM Lotus software wiki program today.

Introduction

Lotus Notes Traveler provides two-way, over-the-air synchronization between Lotus Domino servers and Microsoft Windows Mobile, Nokia S60, and Apple (iPhone and iPod Touch devices). Lotus Notes Traveler synchronizes email, calendar, to-do, address book, and journal data in real time. The Lotus Notes Traveler server runs as a Lotus Domino server task and is installed on a server running Lotus Domino. The Lotus Notes Traveler server task uses the Lotus Domino directory to automatically look up and find Lotus Notes users. You can get more information about Lotus Notes Traveler.


Test setup

For this study, two Lotus Domino servers were used. One server hosted the Lotus Notes Traveler server task, and the other server hosted the user mail database.

Figure 1 shows the testing configuration.

Figure 1. Testing configuration
Testing configuration

The test strategy is to simulate a typical enterprise custom messaging load and then add mobile devices to that environment. The NotesBench workload generators simulate Lotus Notes client activity by sending mail to random users in the test set and causing other activity on the Lotus Domino mail server. The mobile devices are added into this mix using the device simulators, which keep the changes generated by NotesBench synchronized with the client and simulate user actions on the mobile device.


Lotus Notes Traveler 8.5.1 performance

The performances of the Lotus Notes Traveler server and the Lotus Domino mail server were measured using the Lotus Notes Traveler and N8Mail workloads on the Microsoft Windows 2003 64-bit server operating system. The Lotus Notes Traveler server was on a separate server from the Lotus Domino mail server, and it was exercised by test systems that simulated mobile devices. The driver systems ran scripts that simulated remote devices reading and sending mail messages.

Here are the details the Lotus Notes Traveler client script:

There are 64 iterations of this loop with a 15-minute wait between iterations. Each user starts on a random iteration count (first, fifth, and so on), and each user runs for the entire 64 iterations.

Loop start

  • OPEN/READ five Inbox mails (every iteration)
  • Reply to the first mail (every second iteration)
  • SEND one 10-byte mail to one person (every eighth iteration)
  • SEND another 10-byte mail to three people (every eighth iteration)
  • MOVE one mail from the Inbox to the NotesBench folder (every iteration)
  • DELETE two mails (every iteration)

Loop end

In addition to the user scripts, the Lotus Notes Traveler clients listen for auto-synchronized messages from the Lotus Notes Traveler server. If email or calendar data changed on the server as a result of the NotesBench load drivers, then the Lotus Notes Traveler client automatically synchronized the server changes to the device.

All tests for Lotus Notes Traveler 8.5.1 were done for Windows Mobile 5 (WM5) and Windows Mobile 6 (WM6) and for Apple devices. Nokia devices use the same protocol (SyncML) as the Windows Mobile devices and have similar performance characterstics. Apple devices use a different synchronization protocol (Active Sync) and have different performance characteristics than the Windows Mobile and Nokia devices.

For tests for Lotus Notes Traveler 8.5.1, WM5 and WM6 devices were simulated with 40 percent mail formatted as plain text (WM5) and 60 percent mail formatted as rich text (WM6).

We also ran 100 percent Apple device simulation and mixed load with 50 percent Windows Mobile device and 50 percent Apple device load.


Lotus Domino 8.5.1 server workload

The N8Mail workload that was run on the Lotus Domino 8.5.1 mail server simulated users sending mail using a Lotus Notes 8.5.1 client. This workload is part of a NotesBench workload suite that is part of the tools NotesBench and Server.Load (included as an installation option in the administrator client). For more information about the details of the N8Mail workload, see the developerWorks® Lotus article, “IBM Lotus Notes V8 workloads: Taking performance to a new level.”

For these measurements, the N8Mail workload simulated 4000 Lotus Notes clients sending mail to as many as 6000 mail users (4000 plus the number of mobile device users). The additional 2000 mail users (or as otherwise noted) were the Lotus Notes Traveler registered mail users. These measurements simulate a customer environment in which mail users use either a Lotus Notes client or a Lotus Notes Traveler device but not both at the same time.

Disclaimer

All our tests are lab measurements; therefore, the results might differ from those of the real customer environments. For assistance with capacity planning, consult with the IBM TechLine team, which does capacity planning for new hardware.


IBM Lotus Notes Traveler 64-bit server configuration and results

Lotus Notes Traveler 8.5.1 server (64-bit) was set up as a single Lotus Domino partition server on an IBM eServer™ xSeries® 3550 running Windows 2003 64-bit Enterprise Server with four dual-core processors and 8 GB of physical memory. Network access was through a single 1 GBit Ethernet adapter running in full-duplex mode.

Tables 1 and 2 show the hardware and software configurations for the Lotus Notes Traveler 64-bit server.

Table 1. Hardware configuration for Lotus Notes Traveler 64-bit server
ModelIBM x3550
Processors for test / speed4 Xeon (dual core) at 3.2 GHz
Installed memory8 GB
Active physical drives2 disks, 73 GB each, 10000 RPM, 320 MB
Active logical volumesN/A
Operating systemMicrosoft Windows 2003 Server Enterprise (64-bit)
Table 2. Software configuration for Lotus Notes Traveler 64-bit server
Notes.ini settingsTo increase the Lotus Notes Traveler initial size of javaheap
Used for all testsNTS_Java_Parms=-Xms625M -Xmx625M
HTTP settingsValue
Persistent connections timeout310 seconds
Output timeout310 seconds
NTSConfig.xml settingsValue
TSS_SYNC_THREADS1.2 times number of devices
WORKER_THREADS ACTIVE_THREADS2 times SYNC threads

1.2 times number of devices
TSS_ADDRESSCACHE_MAX_ENTRIES8000

The Lotus Notes Traveler server must be tuned for the number of active users that it supports. The settings changed for this test are indicated in table 2. See the performance tuning section of the Lotus Notes Traveler 8.5.1 Information Center for more information.

In addition to these tuning parameters, to accommodate more devices on the same server, the following Windows TCP/IP stack changes were needed in the local registry:

  1. [HKEY_LOCAL_MACHINE \System \CurrentControlSet \Services \Tcpip \Parameters]
    MaxUserPort = 15000

    This value was increased from 5000 to 15000 for our test.

  2. [HKEY_LOCAL_MACHINE \System \CurrentControlSet \Services \Tcpip \Parameters]
    MaxFreeTcbs = 10000

    This value was increased from 2000 to 10000 for our test.

  3. [HKEY_LOCAL_MACHINE \System \CurrentControlSet \services \Tcpip \Parameters]
    MaxHashTableSize = 512

    This value was increased from 512 to 2650 for our test.

  4. [HKEY_LOCAL_MACHINE \SYSTEM \CurrentControlSet \Services \Tcpip \Parameters]
    EnableConnectionRateLimiting = 0

Results

Figure 2 shows the processor resource utilization for the Lotus Notes Traveler 8.5.1 64-bit server on 64-bit Windows 2003 at Windows Mobile and Nokia simulated user loads. Figure 3 shows the processor resource utilization for Apple devices and a mixed device workload.

Figure 2. Lotus Notes Traveler 8.5.1 64-bit configuration: Percent processor busy
Lotus Notes Traveler 8.5.1 64-bit configuration: Percent processor busy
Figure 3. Lotus Notes Traveler 8.5.1 64-bit configuration: Percent processor busy
Lotus Notes Traveler 8.5.1 64-bit configuration: Percent processor busy

Table 3 shows the memory usage for different types of loads.

Table 3. Memory usage for different types of loads
Type of testMemory usage in MBComments
2000 Windows Mobile devices1600WM5 and WM6 devices
2000 mixed devices155050 percent WM and 50 percent Apple devices
1500 Apple devices1500N/A

Table 4 shows the mix of WM5 and WM6 users for the tests shown in figure 2.

Table 4. Mix of users for each of the Lotus Notes Traveler user loads
Total number of users Number of WM5 usersNumber of WM6 users
1000400600
20001200800

For the measurements done to obtain the results in figures 2 and 3, the Lotus Domino mail server ran at a steady state of 4000 N8Mail users, and the number of Lotus Notes Traveler users was incremented as shown in the figure. The Lotus Domino mail server was configured for 6000 mail users, out of which 4000 mail users simulated Lotus Notes client traffic using the N8Mail workload. The number of Lotus Notes Traveler users simulating device traffic was then increased from 1000 users to 2000 users. For a mixed load of 2000 users, we used 1000 WM and 1000 Apple users.

Table 5 shows several key resource utilization metrics on the Lotus Notes Traveler server at a load of 2000 active Lotus Notes Traveler Windows Mobile users and 4000 active Lotus Domino N8Mail users.

Table 6 shows similar metrics on the Lotus Notes Traveler server at a load of 1500 active Lotus Notes Traveler Apple device users and 4000 Lotus Domino N8Mail users.

Table 5. Resource usage at 2000 active Lotus Notes Traveler Windows Mobile users and 4000 Lotus Domino N8Mail users
ResourceResults
Processor percent busy46
Total disk reads/second0.15
Total disk writes/second393
Lotus Notes Traveler server virtual memory (Mbytes)1603
Network bytes / second1766
Table 6. Resource usage at 1500 active Lotus Notes Traveler Apple device users and 4000 Lotus Domino N8Mail users
ResourceResults
Processor percent busy51.21
Total disk reads/second0.05
Total disk writes/second275
Lotus Notes Traveler server virtual memory (Mbytes)1498
Network bytes / second3102

Comparison between Lotus Notes Traveler 8.5.1 and Lotus Notes Traveler 8.5

Comparing Lotus Notes Traveler 8.5 and 8.5.1 can be done only based on the Windows Mobile device load. Lotus Notes Traveler 8.5 did not support Apple devices.

Figure 4 shows the processor resource utilization of the Lotus Notes Traveler 8.5.1 64-bit server compared to the same load on Lotus Notes Traveler 8.5 64-bit server on 64-bit Windows 2003.

Figure 4. Processor usage of Lotus Notes Traveler 8.5 and 8.5.1
Processor usage of Lotus Notes Traveler 8.5 and 8.5.1

Figure 5 compares the memory usage of Lotus Notes Traveler 8.5 and 8.5.1 for 1000 and 2000 Windows Mobile devices.

Figure 5. Memory usage of Lotus Notes Traveler 8.5 and Lotus Notes Traveler 8.5.1 for 1000 and 2000 Windows Mobile devices
Memory usage of Lotus Notes Traveler 8.5 and Lotus Notes Traveler 8.5.1 for 1000 and 2000 Windows Mobile devices

Lotus Domino mail server configuration and results

Tables 7 and 8 show the hardware and software configurations for the Lotus Domino mail server.

Table 7. Hardware configuration
ModelIBM x3850
Processors for test / speed4 Xeon (dual core) at 3.0 GHz
Installed memory12 GB
Active physical drives42 disks
Active logical volumes6 arrays of RAID 0
Operating systemWindows 2003 Server Enterprise 64-bit
Table 8. Software configuration
UseNotes.ini settings
Used for all testsserver_pool_tasks=80
server_max_concurrent_trans=100
Show_Server_Performance=1
RouterDbCacheSize=6100
NSF_DBcache_maxentries=6000
SCHEDULE_NO_VALIDATE=1
NSF_Buffer_Pool_Size_MB=250

The mail server runs Lotus Domino 8.5.1 32-bit server set up as a single Lotus Domino partition server on an eServer xSeries 3850 running Windows 2003 64-bit Standard Server, with four dual-core processors and 12 GB of physical memory. The storage configuration was an IBM DS4000® series (FAStT) hardware array with 42 disk drives. The Lotus Domino executable files were installed on one RAID 0 logical volume of 200 GB. The mail databases were spread across three RAID 0 logical volumes, each 800 GB. Mail journaling was set up on a separate logical volume. Network access was through a single 1 GBit Ethernet adapter running in full-duplex mode.

Figure 6 shows the processor resource utilization on the Lotus Domino mail server at various Lotus Notes Traveler Windows Mobile user loads and 4000 N8Mail users. This utilization was measured with the 64-bit Windows 2003 server configuration for Lotus Notes Traveler Server 8.5 64-bit server.

Figure 6. Processor utilization on the Lotus Domino 8.5.1 mail server
Processor utilization on the Lotus Domino 8.5.1 mail server

When increasing the number of Lotus Notes Traveler Windows Mobile users from 1000 to 2000, we measured a 4 percent relative processor usage increase on the Lotus Domino mail server that directly correlated with the load we had on the Lotus Notes Traveler server. The processor overhead of enabling Lotus Notes Traveler users on the Lotus Domino mail server is moderate because of the efficient use of polling for Lotus Domino mail by the Lotus Notes Traveler server.


Conclusion

The performance measurements of the Lotus Notes Traveler 8.5.1 server and the Lotus Domino 8.5.1 mail server in a remote configuration show that a Lotus Notes Traveler 8.5.1 64-bit server was able to support 2000 Windows Mobile users or 1500 Apple device users, or as many as 2000 mixed users for the workloads presented in this report.

The Lotus Notes Traveler 64-bit server, with four dual-core processors and 8 GB memory, was at approximately 46 percent processor utilization and at 1.6 GB memory usage for 2000 Windows Mobile devices active users. For 1500 Apple devices active users, the processor utilization was about 51 percent and 1.5 GB memory usage.

With the Lotus Domino 8.5.1 mail server at steady state of 4000 N8Mail users and 2000 (64-bit) Lotus Notes Traveler Windows Mobile active users, processor utilization on the Lotus Domino mail server increased moderately.

To support more than 2000 active Lotus Notes Traveler users, customers need to deploy more than one Lotus Notes Traveler 8.5.1 64-bit server. The processor utilization on the Lotus Domino mail server can be monitored to ensure that the Lotus Domino mail server can handle the added activity from the Lotus Notes Traveler users at peak load.

Note that the Lotus Notes Traveler server is a memory-intensive application. Deploy Lotus Domino 64-bit servers to support large numbers of mobile device users.

Appendix 1. Notes.ini settings

Table 1. Hardware configuration for Lotus Notes Traveler 64-bit server
Notes.Ini parameter nameExplanation
Create_R8_DatabasesApplies to Lotus Domino 8 servers. Set it to 1 to create ODS48 databases on Lotus Domino 8 servers.
NSF_Buffer_Pool_Size_MBSpecifies the maximum size in bytes of the NSF buffer pool, a section of memory dedicated to buffering I/O transfers between Lotus Domino and disk storage.
NSF_Dbcache_MaxentriesSpecifies the number of databases that a server can hold in its database cache at one time.
RouterDbCacheSizeSpecifies the size of the router database cache used to cache mail databases in the router process.
SCHEDULE_NO_VALIDATEEnables or disables whether SchedMgr validates its busytime database entry on a daily basis, as follows:
0 (default) - Enables validation
1 - Disables validation
Server_MAX_CONCURRENT_TRANSSets the limit for the number of concurrently scheduled transactions on a server.
Server_Pool_TasksSpecifies the total number of physical threads in the Lotus Domino server (DbServer) threadpool.
Server_Show_PerformanceSpecifies whether server performance events are displayed on the console. Set this variable to 1 to display server performance events on the console.

Resources

Comments

developerWorks: Sign in

Required fields are indicated with an asterisk (*).


Need an IBM ID?
Forgot your IBM ID?


Forgot your password?
Change your password

By clicking Submit, you agree to the developerWorks terms of use.

 


The first time you sign into developerWorks, a profile is created for you. Information in your profile (your name, country/region, and company name) is displayed to the public and will accompany any content you post, unless you opt to hide your company name. You may update your IBM account at any time.

All information submitted is secure.

Choose your display name



The first time you sign in to developerWorks, a profile is created for you, so you need to choose a display name. Your display name accompanies the content you post on developerWorks.

Please choose a display name between 3-31 characters. Your display name must be unique in the developerWorks community and should not be your email address for privacy reasons.

Required fields are indicated with an asterisk (*).

(Must be between 3 – 31 characters.)

By clicking Submit, you agree to the developerWorks terms of use.

 


All information submitted is secure.

Dig deeper into IBM collaboration and social software on developerWorks


static.content.url=http://www.ibm.com/developerworks/js/artrating/
SITE_ID=1
Zone=Lotus
ArticleID=472340
ArticleTitle=IBM Lotus Notes Traveler 8.5.1 performance
publish-date=03092010