A look at MOSS 2007 Performance Part I – Performance Counters
Introduction
This series is intended to be a multi-part look at MOSS performance. There are many articles out there related to the performance analysis of MOSS but I wanted to attempt to compile some of that information and add some personal insight to the pool.
The scenario this is to address is to attempt the first steps in analyzing the question of "Why is my MOSS performing in a less ideal way that I would like?". We are also going to take a look at some ways in which we can proactively monitor the system with no additional tools to purchase.
Part one of this series will focus on Performance Counters.
Performance Counters
Assumption: Windows Server 2003 or higher is being utilized as the host operating system
General Counters
The following table will outline the performance counters for the base operating system that should be monitored along with thresholds and what the tolerance factor may mean if we are outside of those thresholds.
CPU
Counter |
Threshold |
General Recommendation |
Description |
Processor\% Processor Time\_Total |
>80%-85% over 3 intervals |
Add CPUs if utilization is consistently High |
The total percentage of processor usage for a server. |
System\Processor Queue Length\(N/A) |
>#CPU's + 1 |
Add Faster CPU's |
Exceeding the threshold indicates that the processors are not fast enough to process requests as they occur and that you are running out of processor power. |
Processor\Interrupts\Sec |
>1000 |
May indicate Hardware Issue |
|
Memory
Counter |
Threshold |
General Recommendation |
Description |
Memory\Pages/sec\(N\A) |
>200 |
Increase RAM |
|
Memory\Pages Faults/sec\(N\A) |
> (Drives x 4096) |
Increase RAM |
A high rate of page faults indicates a lack of physical memory. |
Memory\Available Mbytes\(N\A) |
<10% Available |
Increase RAM |
The amount of physical memory in MBs, immediately available for allocation to a process or for system use on the server. |
Memory\Pool Nonpaged Bytes\(N\A) |
<10% |
Increase RAM |
|
PagingFile\%Usage |
>70% |
Faster Disks/Better Controller/Add Disks/Increase RAM |
Review this value in conjunction with memory—available megabytes and page faults per second—to understand paging activity on the server. |
Disk
Counter |
Threshold |
General Recommendation |
Description |
PhysicalDisk\% Disk Time\DriveLetter |
>80% |
Controller or RAM |
|
PhysicalDisk\Current Disk Queue Length\DriveLetter |
>2 |
Faster Disks/Better Controller/Add Disks |
|
LogicalDisk\%Idle Time\DriveLetter |
>20% |
Faster Disks/Better Controller/Add Disks |
Reports the percentage of time during the sample interval that the disk was idle. If this value is very low, the logical disk is very busy. |
Network
Counter |
Threshold |
General Recommendation |
Description |
Network interface\BytesTotalPer2ndInterface |
>50% Avail Network Speed |
Add Faster NICs/Check Switch |
The rate at which bytes are sent and received over each network adapter. |
ASP.NET
Counter |
Threshold |
General Recommendation |
Description |
ASP.NET\Requests per second_total |
|
This is a great monitor for capacity as we will look at in a different part of this series |
The rate at which bytes are sent and received over each network adapter. |
MOSS WFE
Object |
Counter |
Threshold |
Description |
Web service* |
Get requests per second_total or individual portal or per Windows SharePoint Portal Services site |
Through ongoing monitoring, trends begin to emerge that equate get requests per second with CPU consumption. |
Generally speaking, this is the rate at which clients are requesting information from the front-end Web servers. |
MOSS Query Server
Object |
Counter |
Threshold |
Description |
Search |
Query rate |
200 per second |
The number of queries posted to the server per second. Keep in mind this counter should be watched on the dedicated search servers. |
Search |
Succeeded queries |
This counter should be used mostly for troubleshooting search problems. |
The number of queries that produce successful searches; monitor this counter along with the failed queries counter when troubleshooting search problems. |
Search catalogs |
Number of documents |
Microsoft has tested up to 5,000,000 documents. |
This counter identifies the total number of documents in the catalog. |
Search catalogs |
Queries rate_(index names or all instances) |
Indicates which indexes are most searched by users. |
The number of queries posted to indexes per second; in conjunction with other performance data, this counter can help determine if your index configuration can be optimized. |
Web service* |
Get requests per second_total or individual portal or Windows SharePoint Portal Services site. |
With ongoing monitoring, trends emerge that equate requests per second with CPU consumption. |
Generally speaking, this is the rate that clients are requesting information from the front-end Web servers. |
MOSS Index Server
Object |
Counter |
Threshold |
Description |
Search indexer catalogs |
Number of documents |
Microsoft has tested up to 5,000,000 documents. |
The number of documents in the catalog |
Search indexer catalogs |
Index size MB |
256,000 (256 GB) |
The current size of index data in megabytes |
I want to thank a colleague of mine, Kirk for providing much of this information that was MOSS specific
References:
English, B. (2004). Microsoft SharePoint Products and Technologies Resource Kit. New York: Microsoft P.