Server Performance

HCL Domino Domino Forum
1

Hi everyone,

I wanted to see if I could get some pointers on what I can do to optimize the performance on my mail server. This mail server is currently in a cluster an house very large mail databases. Every so often the server restarts itself and throws errors on the console stating “insufficient system resources to perform task”. This server currently has 16 processors and 15gb of memory, so you would think that it has enough resources to complete task. Any insight on this matter would be gladly appreciated. Thanks

2

Subject: Server Performance

A TechNote that is a bit dated:

http://www-1.ibm.com/support/docview.wss?uid=swg21085230

Knowledge Collection on Server Performance:

http://www-1.ibm.com/support/docview.wss?uid=swg27003037

3

Subject: Server Performance

What operating System???

4

Subject: RE: Server Performance

You said : Every so often the server restarts itself

Does it crash?

If so , any NSD? Can you post the FATAL THREAD section of the NSD files with the few lines under it

For example:

############################################################

FATAL THREAD 1/2 [ domdsm:107a8:67076]

FP=0x0012a34c, PC=0x60069a86, SP=0x0012a2f0, stksize=92

EAX=0x01a3088c, EBX=0x01a48650, ECX=0x0000bffe, EDX=0x01a325e6

ESI=0x00004000, EDI=0x0032ff78, CS=0x0000001b, SS=0x00000023

DS=0x00000023, ES=0x00000023, FS=0x0000003b, GS=0x00000000 Flags=0x00010206

Exception code: c0000005 (ACCESS_VIOLATION)

############################################################

@[ 1] 0x60069a86 nnotes._fdDelete@4+118 (4000,12a368,608770ef,4000)

@[ 2] 0x60069a0c nnotes._OSFileClose@4+12 (4000,12adc0,12a380,60872ea7)

@[ 3] 0x608770ef nnotes._sqloclose+15 (4000,12b59c,5010,0)

@[ 4] 0x60872ea7 nnotes._sqlpgclf+71 (12adc0,1aaa2cb,70,1aaa2f3)

@[ 5] 0x60885458 nnotes._sqlpgfdl+616 (12b9a4,1aaa2cb,41,12be44)

@[ 6] 0x60876a6f nnotes._sqlpgcdl+927 (12bf44,12b9a4,0,12be44).

Date

Topic

Frequently asked question - How to analyze Notes/Domino NSDs? (Jean-Yves Riverin)

or call IBM support.

JYR

5

Subject: RE: Server Performance

Thank everyone for your insight on this. We have looked into the best practices from Lotus and performed as many tweaks they we could, but users still complain about accessing mail is slow. We have tried to send out several notices about inbox management. The server doesn’t actually crash were it generates a nsd. It will restart itself and than the restart analysis runs and recovery manager. We are currently trying to push the upgrade to 7 to upper management. We are also trying to convince management that we need to look into setting up some type of archiving to get mail file sizes down.

What are the largest mail files that others have to deal with? Our mail files range from 14mb to 17gb, we currently have about 5000 users and about half of users mail files are over 500mb.

Once again thanks for everyones post.

6

Subject: RE: Server Performance

I’ve seen mail files over 20 GB. Sure can’t recommend it! :wink:

Why push an upgrade just to 7? 8 has significant improvements in server performance. It also has an inbox management agent that will automatically take older messages out of the inbox, leaving them accessible through All Documents. Even if you can’t upgrade clients to 8, you can still get some pretty significant benefits.

And 8.5, when released, will have new features that will further improve performance and storage management.

7

Subject: RE: Server Performance

I really wish we could migrate to 8. Its like pulling teeth to get management to agree upon migrating to ver 7.

8

Subject: RE: Server Performance

I believe you when you say that getting your management to consider upgrading to 7 is like pulling teeth, because management’s natural inertia is very strong; but why would getting them to consider 8 be even harder? Why even consider doing an upgrade that still leaves you without major performance and stability improvements that are in 8?

Note that I’m talking about servers, not clients. I understand why there could be impediments to migrating clients to 8, and a migration of clients to 7 would bring benefits, so a customer still on 6.x might consider migrating clients to 7. But I don’t see that logic applying to servers at all. 7 is up to 7.0.3 FP1. Counting all the point releases and fix packs, there have been 7 releases in the 7.x series, over a period of 3 years; but 8.0.1 is out, so the old “we don’t upgrade to point-zero releases” objection is gone. Does 8 have to get through a few more point releases and fix packs before it can be considered for server upgradess? If that’s not it, then what are the impediments to upgrading servers to 8 that are not there with upgrading to 7.0.3?

BTW: Just trying to understand the way of thinking here. Not intending to give you a hard time personally. I’ve been doing this stuff a long time, and I’m used to the reluctance to upgrade to the latest point-zero release. But I’m used to that reluctance fading quickly after the point-one release. IBM did a long beta for 8, and 8.0 has had nearly a year in the field. 8.0.1 is out, and 8.0.2 is in the works. 8.5 is in public beta. It seems to me that IBM has done things the right way to build confidence that 8 is a stable release, so it really surprises me to see that the reluctance is so strong that there are customers still considering upgrading from 6.x to 7 but not to 8.

9

Subject: Server Performance

First of all, you will probably gain far more from upgrading from 6.5.2 to 7, or better yet 8, than you will from any tweaks you might make. There have been dramatic improvements in server performance in recent releases.

Also, is your server partitioned? Due to shared memory limits, there’s no way that a single non-partitioned Domino server could ever take advantage of all that RAM.

10

Subject: Server Performance

Hi,

Best practices for large Lotus Notes mail files

From Ed Brill’s blog Inbox management

Notes/Domino Best Practices: Performance

IBM - Administrator Guide for Domino Server maintenance

Abstract The purpose of this document is to describe when to use the Lotus Domino utilities (Updall, Compact, Fixup) and to explain how to use them for normal and abnormal maintenance.

JYR

11

Subject: Server Performance

Tuning Memory Usage on Your IBM Lotus Domino Serverhttp://www-10.lotus.com/ldd/sandbox.nsf/ecc552f1ab6e46e4852568a90055c4cd/cb533a62695b1f08852571370053fa43?OpenDocument&Highlight=0,id111

You said that you have a cluster. Are you mail dbs in synch(Same number of docs)?

How many cluster replicator do you have?

Issue this command on both servers of your cluster:

show stat replica.cluster*

It will give you something similar:

Replica.Cluster.Docs.Added = 5022

Replica.Cluster.Docs.Deleted = 62226

Replica.Cluster.Docs.Updated = 162746

Replica.Cluster.Failed = 4149

Replica.Cluster.Files.Local = 1057

Replica.Cluster.Files.Remote = 2066

Replica.Cluster.Retry.Skipped = 592538 <–check this

Replica.Cluster.Retry.Waiting = 0

Replica.Cluster.SecondsOnQueue = 0 <–check this

Replica.Cluster.SecondsOnQueue.Avg = 18 <–check this

Replica.Cluster.SecondsOnQueue.Max = 1894 <–check this

Replica.Cluster.Servers = 2

Replica.Cluster.SessionBytes.In = 378840196

Replica.Cluster.SessionBytes.Out = 1747925166

Replica.Cluster.Successful = 433515

Replica.Cluster.WorkQueueDepth = 0 <–check this

Replica.Cluster.WorkQueueDepth.Avg = 7 <–check this

Replica.Cluster.WorkQueueDepth.Max = 363 <–check this

If you run good servers, you can add up to 6 cluster replicators without any problems.

JYR

12

Subject: Server Performance

You can try compressing the attachments that are inside these email DBs thus gaining storage, bandwidth, and performance.Take a look at zipmail (www.mknetwork.com)

HTH