What do I have to do to have the values for CPU Utilization appear in the Domino Administrator (Server, Status, Server Tasks) - Far Right Column. I cannot seem to find any mention of it in the Administration Help… or in these postings…
Thanks,
C>
Subject: CPU Utilization
Do you get 0% values in here in nothing at all? My 6.5 Admin client has 0% most of the time against all the tasks on our 6.0.1 CF1 and 5.0.11 servers (but blanks for Database Server entries and Port entries).
Launch a nice processor intensive task (such as LOAD UPDALL db.nsf -R) and press F9 and the CPU utilization jumps as you would expect.
If you’re really getting nothing, let us know what version of the admin client and server you are using.
Rupert Clayton
London
Subject: Nothing at all…
6.0.2 CF 2 Administration Client against a 6.0.2 CF 2 Domino Server… the entire column is blank.
Thanks,
C>
Subject: RE: Nothing at all…
I’ve tried a 6.0.1 CF1 Admin client against a 6.0.2 CF1 server (admittedly, not your combination), and the utilisation appears fine.
So it could still be a bug in your particular client or server version…
But here’s another line of inquiry. Do you have platform statistics enabled. I believe it should now be standard in 6.0.x and later 5.0.x releases. But I guess it could be turned off. There’s some info in this article on server health monitoring (but platform stats are actually independent of health monitoring).
The posting say you’re running Windows 2003. Another possibility is that Domino somehow can’t get CPU utilisation on 2003 Server.
Yes, I am kind of grasping at straws now!
Rupert
Subject: Me too…
…I have tried turing on all the stats and making sure the registry entries exits for permon / notestat but so far no change.
I guess it might have to be a support call.
Thanks,
C>
Subject: Possible CPU utilization stats solution
Alright, some more clues.
I have one 5.0.11 server on which I can see CPU utilization and another on which I can’t.
The first server also responds to SH STAT PLATFORM with a list of platform statistics. The other one responds with “PLATFORM not found in stat table”
That suggests to me it’s definitely related to enabling platform statistics. (The machine that works is on NT, the one that doesn’t is on 2000, but that doesn’t seem a likely cause.)
There is the following Notes.ini variable
PLATFORM_STATISTICS_DISABLED=1
but I’m guessing that’s not set on your system.
So you (and I) need to look at why platform stats aren’t running.
Domino Admin help has the following (not very helpful) comment:
"Platform statistics take a few minutes to initialize upon Domino server startup. If you issue a Show Stat Platform command before initialization, you get the following message:
Platform not in statistics table
Wait a few minutes and then issue the Show Stat Platform command again."
[…several minutes of investigation later]
OK, looks like there’s also a NOTES.INI variable
PLATFORM_STATISTICS_ENABLED=1
(God only knows what happens if you have both of these set.)
PLATFORM_STATISTICS_ENABLED=1 is present on my functioning servers and not on my non-reporting servers. So that might do the trick. I’ll try it and report back.
HTH,
Rupert
Subject: Hmmm…
…well I added the INI parameter and recycled, and I don’t get the ‘PLATFORM not found…’ message any more when issuing a ‘sh stat platform’ command:
sh stat PLATFORM
®
sh stat PLATFORM
rectory Manager
sh stat PLATFORM
,
sh stat PLATFORM
sh stat PLATFORM
6/O=Company Name Here
sh stat PLATFORM
®
sh stat PLATFORM
rectory Manager
hmmm! This is on 2003 - so perhaps the changed the structure from NT to 2000 & 2003.
C>
Subject: From 5.0.2 release notes
I tried the ‘platform time 5’ & ‘platform reset’ commands suggested below too. Nothing as yet…
Server
Performance Monitoring Statistics (NT Intel and Solaris Sparc only)
Domino will track performance metrics of the operating system and output the results to the server. Type “show stats platform” at the server console to display them. This feature is disabled by default in R5.0.2. You can enable it by setting the parameter:
PLATFORM_STATISTICS_ENABLED=1
…in the NOTES.INI file and restarting your server. On Windows NT, the following registry setting:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\notestat\Performance\Library
…must be set to NNOTES.DLL in your executable directory. Otherwise performance monitoring will not run. Also on Windows NT, diskperf needs to be enabled for disk metrics to be captured.
Metrics are collected continuously, unless disabled or reset (see commands below), or unless the server is brought down. The server resets this information each time it is started. This feature also expects certain Notes processes to be present. Performance counters for absent Notes processes will have zero values. This is normal behavior.
The server monitors the following metrics:
Performance Metric Description
Platform.Memory.KBFree The amount in kilobytes of virtual memory available on the system. This is equivalent to Memory/AvailableBytes in Windows NT Performance Monitor divided by 1024.
Platform.Memory.PagesPerSec The number of pages read to or written from the disk. Equivalent to Memory/Pages/second in NT Perfmon.
Platform.System.TotalUtil The average percent CPU utilization of all CPU’s on the system. Equivalent to System/% Total Processor Time in NT Perfmon.
Platform.System.TotalPrivUtil The average percent CPU utilization in privileged mode of all CPU’s on the system. Equivalent to System/% Total Privileged Time in NT Perfmon.
Platform.System.TotalUserUtil The average percent CPU utilization in user mode of all CPU’s on the system. Equivalent to System/% Total User Time in NT Perfmon.
Platform.LogicalDisk._Total.1._Total.1.PctTime The percentage of the sampling interval that all logical disks on all physical disks are servicing read or write requests. Equivalent to LogicalDisk/% Disk Time in NT Perfmon. As in NT Perfmon, the first “_Total” refers to all physical disks, and the second “_Total” refers to all logical disks.
Platform.LogicalDisk._Total.1._Total.1.AvgQueueLength (NT only) The average number of both read and write requests that were queued for all logical disks on all physical disks during the sample interval. Equivalent to LogicalDisk/Avg. Disk Queue Length in NT Perfmon.
Platform.LogicalDisk._Total.1._Total.1.ServiceTime (Solaris only) The amount of time in milliseconds for a read or write command to complete.
Platform.Process.nserver.1.Util (NT only) The CPU utilization of the first server process. This is the percentage of the sample interval that the process used the CPU. Equivalent to Process/% Processor Time in NT Perfmon.
Platform.Process.nrouter.1.Util (NT only) The CPU utilization of the first router process.
Platform.Process.nreplica.1.Util (NT only) The CPU utilization of the first replica process.
Platform.Process.nreplica.2.Util (NT only) The CPU utilization of the second replica process.
Platform.Process.nreplica.3.Util (NT only) The CPU utilization of the third replica process.
Platform.Process.nreplica.4.Util (NT only) The CPU utilization of the fourth replica process.
Platform.Process.nupdate.1.Util (NT only) The CPU utilization of the first update process.
Platform.Process.nupdate.2.Util (NT only) The CPU utilization of the second update process.
Platform.Process.nupdate.3.Util (NT only) The CPU utilization of the third update process.
Platform.Process.nupdate.4.Util (NT only) The CPU utilization of the fourth update process.
Platform.Process.nadminp.1.Util (NT only) The CPU utilization of the first adminp process.
Platform.Process.nadminp.2.Util (NT only) The CPU utilization of the second adminp process.
Platform.Process.nadminp.3.Util (NT only) The CPU utilization of the third adminp process.
Platform.Process.nadminp.4.Util (NT only) The CPU utilization of the fourth adminp process.
Platform.Process.nhttp.1.Util (NT only) The CPU utilization of the first http process.
For non-process metrics (those not starting with “Platform.Process”), average, minimum, and maximum statistics are also reported. For example, in addition to Platform.Memory.KBFree, there are the following secondary statistics:
Platform.Memory.KBFree.avg The average of all KBFree samplings for a session. This is calculated by summing all the KBFree samplings, and dividing by the number of samplings.
Platform.Memory.KBFree.min The smallest KBFree sampling for a session
Platform.Memory.KBFree.max The largest KBFree sampling for a session
If a process is removed, the metric for that process will go to 0.
The PLATFORM command allows you to control this feature at the console. It has the following form:
PLATFORM
These are the main arguments:
Argument(s) Description
TIME With an optional argument, changes the sampling rate to the specified value in minutes. Otherwise displays the current sampling rate. The default is 1 minute.
RESET Initiates the beginning of a new statistics monitoring session; resets secondary metrics.
WAIT Pauses the collection of performance data
RUN Resumes the collection of performance data
For example, to start a new performance data monitoring session with a sampling rate of 5 minutes, type the following commands at the console:
platform time 5 - The server collects performance data every 5 minutes.
platform reset - Statistic values sampled before this command was issued are not used in calculating average, maximum, or minimum.
Changes have also been made to the Stats and Events database so these statistics can be treated like any other stat (i.e. thresholds, etc. They can be viewed through the Admin client.
Subject: Fixed with upgrade to 6.0.3