Server Performance Metrics

With up.time's graphical server monitoring software, you can instantly graph and visually analyze all critical server resources within your enterprise regardless of what operating system you are running, including: CPU, Memory, Disk, Processes, Workload, Network, User, Services Status and Configuration Data.

You can quickly determine how critical resources are being consumed over the past few minutes, hours, days, weeks or months. With the data collected by an up.time server monitor, you can immediately graph all server resource statistics, server availability statistics, generate ad-hoc management reports or schedule resource reports for automatic generation and delivery to your email inbox.

Using up.time's Server Agents and agent based monitoring, you get the most comprehensive, in-depth capture of operating system specific server resource and performance metrics. Simply install the specific up.time Server Agent on each operating system you need to monitor and you will get over 100 different metrics to include in your graphs and reports.

Alternatively, servers can be monitored agentless by ensuring they have the latest version of Net-SNMP installed (Download at: http://net-snmp.sourceforge.net/). If this approach is preferred only the up.time Monitoring Station needs to be installed without the up.time Server Agents. However, agentless monitoring will provide a much more limited set of metrics (as available from Net-SNMP) for graphing and reporting purposes. For environments requiring in-depth enterprise server reporting using up.time agents is highly recommended.

Note some variances in metrics listed below do exist between Windows, UNIX, Linux and Novell NRM systems. All of the statistics obtained via Net-SNMP are identical for every operating system where Net-SNMP is installed.

CPU

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
% USR (User) * *
% SYS (System) * *
% WIO (Wait I/O) * *
% TOTAL * *
Run Queue Length   *
Run Queue Occupancy   *

Multi-CPU

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
CPU ID * *
USR * *
SYS * *
WIO * *
System Time Per CPU   *
Wait I/O Time Per CPU   *
User + System + Wait I/O   *
Spins on Mutexes per CPU   *
Interprocess Cross-Calls   *
Interrupts (INTR)   *

Memory

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Free Memory * *
Cache Hit Rate   *
PageOut Per Second   *
PageIn Per Second   *
PageFree Per Second   *
PageScan Per Second   *
Free Swap * *

Process - Per Process

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Process Name * *
PID * *
PPID   *
UID   *
GID   *
Memory Consumed * *
RSS   *
CPU % Utilization by Process   *
Memory % Utilization by Process   *
Process Start Time   *
Process Run Time *
(in centi-seconds on CPU)		 *

Process - System Level

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Number of Processes Running * *
Number of Blocked Processes   *
Number of Waiting Processes   *
Execs Per Second   *
Process Creation Rate   *

Process - Novell NRM Only

Metric Agentless
(Net-SNMP)		 Novell NRM
(no agent required)
DS Thread Usage   *
Allocated Server Processes   *
Available Server Processes   *
Abended Thread Count   *
Available Memory   *
Work To Do Response Time   *
Packet Receive Buffers   *
Available ECBs   *
LAN Traffic   *
Available Disk Space   *
Disk Throughput   *
Connection Usage   *

Workload

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Workload by Process   *
Workload by User   *
Workload by Group   *
Workload Top 10 by Process   *
Workload Top 10 by User   *
Workload Top 10 by Group   *

Network

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Network I/O * *
In Kbps *
(octets)		 *
Out Kbps *
(octets)		 *
In Errors *
(octets)		 *
Out Errors *
(octets)		 *
Collisions   *
TCP Retransmits   *

Disk - Physical/Logical Disk Performance

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Disk (Spindle) Name   *
Usage (% Busy)   *
Throughput (Blk/s)   *
Read/Writes/s   *
Average Queue Length   *
Average Service Time   *
Average Wait Time   *

File System Capacity:

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
File System Name * *
File System Size * *
File System Used * *
File System Available   *
File System % Used   *

Veritas Volume Manager

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
DG/Volume/Subdisk   *
I/O Operations (Read/Write)   *
Block Throughput (Read/Write)   *
Average Service Time (Read/Write)   *

User

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Number of Users Logged In * *
User Name + Number User Sessions   *

Services

Metric Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
Service Status * *
Last Check Time * *
Duration in Current State * *
Acknowledged * *
Current State * *
Time in Okay State * *
Time in Warning State * *
Time in Critical State * *
Time in Unknown State * *
% Time Okay * *
% Time Warning * *
% Time Critical * *
% Time Unknown * *
Outage Diagnosis * *

Configuration

Configuration Item Agentless
(Net-SNMP)		 Agent-Based
(up.time Agent)
System Name * *
Domain (YP)   *
Architecture * *
Operating System Version   *
Number of CPUs * *
Memory Size * *
Page Size   *
Swap Size * *

Logical Volume Manager Installed?
(Veritas or Solaris DiskSuite)

   *
Host ID   *
CPU Configurations *
(Text only and non-standard across platforms)		 *
(Detailed by cpu_id, speed, module, board, etc).
Network Interface Configurations * *
File System Layout * *
Volume Manager Layout   *

WebLogic Performance Metrics

Metric Agentless Agent-Based
(up.time Agent)
Failures To Reconnect *  
Connection Delay Time *  
Active Connections Current Count *  
Active Connections High Count *  
Leaked Connections Count *  
Current Capacity *  
Number of Available Sessions *  
Waiting For Connection Current Count *  
Access Total Count *  
Beans In Current Use Count *  
Cached Beans Current Count *  
Activation Count *  
Passivation Count *  
Cache Access Count *  
Cache Miss Count *  
Heap Size Current *  
Heap Free Current *  
Open Sockets Current Count *  
Accept Backlog *  
Execute Thread Current Idle Count *  
Pending Request Current Count *  
Transaction Committed Total Count *  
Transaction Rolled Back Total Count *  
Invocation Total Count *  

WebSphere Performance Counters

Counter Agentless Agent-Based
(up.time Agent)
Create Count *  
Remove Count *  
Passivate Count *  
Method Call Count *  
Method Response Time *  
Connection Pool Size *  
Free Pool Size *  
Percent Used *  
Wait Time *  
Close Count *  
Connection Pool Count *  
Waiting Thread Count *  
Use Time *  
CPU Usage *  
Heap Size *  
Used Memory *  
Active Count *  
Committed Count *  
Rolled Back Count *  
Live Count *  
Time Since Last Activated *  
Servlet Pool Size *  

VMware ESX Metrics

Counter Agentless Agent-Based
(up.time Agent)
CPU *  
Memory *  
Network I/O *  
Disk I/O *  
% Ready *  
% Used *  
% Wait *  
Memory Balloon/Target *  
Memory Overhead *  
Memory Swap In/Out *  
Memory Swap Used/Target *  
Memory Zero *  
Disk Latency (Total, Kernel, Device) *  
Disk Commands Aborted / Issued *  
Disk Bus Resets *  

AIX LPAR Metrics

Counter Agentless Agent-Based
(up.time Agent)
lpar_id   *
putt   *
entitlement   *
cpus   *
capped   *
smt   *
memory_total   *
memory_used   *
net_packets_tx   *
net_packets_rx   *
net_bytes_tx   *
net_bytes_rx   *
disk_kbytes_out   *
disk_kbytes_in   *