308 lines
9.5 KiB
Plaintext
308 lines
9.5 KiB
Plaintext
English French Notes Complete/Exclude
|
|
BERNSTEIN RESPONSE TIME STATISTKICS
|
|
Mean
|
|
Site Avg. From Service Level Data:
|
|
Average of Sites Running BRTM, V2:
|
|
% of total events less than 1 sec:
|
|
% of total events less than 2 sec:
|
|
[Estimated Response Times by Site using Service Level Data]
|
|
[Response Times by Site for BRTM, Version 2]
|
|
[Average Number of Events/Second]
|
|
For node:
|
|
BRTL HOST FILE
|
|
Command
|
|
Included
|
|
Earliest
|
|
Latest
|
|
Best
|
|
Worst
|
|
or over
|
|
XUCM SITE PROFILE
|
|
8986.095/CMP PROFILE FROM
|
|
SITE PROFILE UPDATE
|
|
The MINIMUM and MAXIMUM values are normally used for scaling.
|
|
You can optionally change these values.
|
|
ENTER MINIMUM: //
|
|
ENTER MAXIMUM: //
|
|
THE LOW VALUE MUST BE GREATER THAN THE HIGH VALUE.
|
|
SUBMIT/LOG/NOPRINT/QUE=
|
|
MISC_
|
|
Main Memory
|
|
HARDWARE TYPE
|
|
device type
|
|
device type
|
|
Total blocks
|
|
Free blocks
|
|
<DISK WARNING>
|
|
/ VMS Blocks FREE:
|
|
<VOL SET WARNING>
|
|
/ DSM Blocks FREE:
|
|
HARDWARE PERFORMANCE SUMMARY
|
|
DISK USAGE
|
|
Drive / Type
|
|
IO Rate
|
|
Free
|
|
Time*
|
|
*Disk Response Times are estimated: RT (ms)=(QUE LENGTH/IO RATE)*1000
|
|
Purging 8986.6
|
|
Purging 8986.51
|
|
Purging 8986.5
|
|
Purging 8986.098
|
|
8986.6 UPDATE FROM
|
|
Enter number of days to keep RAW data:
|
|
Enter number of days to keep DAILY AVERAGES:
|
|
Enter number of days to keep RAW RESPONSE TIME data:
|
|
MANUAL PURGE OF CMP DATA
|
|
Request queued to task number:
|
|
Please enter a positive number.
|
|
DATABASE UTILIZATION
|
|
DSM Blks
|
|
Total Mb
|
|
RTHIST SUMMARY
|
|
ON NODE
|
|
High(Low)
|
|
2SD Range
|
|
System Performance Summary for 8am - 4:30pm,
|
|
(Hourly collections of 15 minutes duration, 10 second sample intervals)
|
|
SYSTEM SUMMARY
|
|
For Node:
|
|
Metric
|
|
2sd Range
|
|
CPU MODES
|
|
COMPUTE QUEUES
|
|
MEMORY/IO
|
|
ALL EVENTS OUTSIDE 3 STANDARD DEVIATIONS ON NODE
|
|
3sd Limits
|
|
I will print your current metric references and the new
|
|
ranges. You will be able to exit this routine if the new ranges
|
|
are unsatisfactory.
|
|
Print current references to which DEVICE?
|
|
>>Current entries in CM METRICS FILE<<
|
|
Computing new references. One Moment...
|
|
Print NEW references on which DEVICE?
|
|
New hardware reference ranges
|
|
No updating occurred.
|
|
OK to file this data
|
|
Updating NOW. One moment...
|
|
Your site parameters indicate that your new data will be overwritten
|
|
the next time the Morning Summary is generated.
|
|
To prevent this, you may delete the entry in the following field.
|
|
No data available!
|
|
Hardware
|
|
Press <RET> to Continue...
|
|
Analysis of which DATE?
|
|
PERFORMANCE ASSESSMENT
|
|
PERFORMANCE ASSESSMENT ON NODE
|
|
Range Unavailable
|
|
PERFORMANCE ANALYSIS FOR DATE
|
|
HW TYPE
|
|
LOCAL RANGE (2 SD)
|
|
PERFORMANCE ASSESSMENT OF DISK DRIVES
|
|
Overall fault rate:
|
|
Maximum fault rate tolerated:
|
|
Queue Length (COM+COMO):
|
|
<<BEGINNING VPM POST-INIT>>
|
|
This post-init will allow you to review and update your
|
|
VPM site file entries. Taskman will also install new VMS com files
|
|
in your VPM host directory.
|
|
TASKMAN MUST BE RUNNING FROM A DCL CONTEXT TO COMPLETE THIS STEP.
|
|
YOU CAN '^' OUT NOW IF THIS IS NOT THE CASE.
|
|
Note that VPM also requires a HFS device, SYS$INPUT, and XUCM RESOURCE.
|
|
IF YOU EXIT NOW, RUN THE OPTION, 'SETUP PERFORMANCE MONITOR'
|
|
LATER.
|
|
The routine ^XUCMTM was written to assist with setting up Taskman
|
|
to run from DCL. Make sure you have SYSPRV and OPER before attempting
|
|
to run this routine.
|
|
...Checking environment.
|
|
Kernel 8.0 required.
|
|
XU*8.0*35
|
|
Kernel Patch, XU*8.0*35 not detected. Please install this patch before proceeding.
|
|
VPM appears to be running and needs to be disabled.
|
|
Deleting file 8986.098 and removing Bernstein RT Data.
|
|
Adding metrics, RTM, and RSP.
|
|
VPM Response Time
|
|
Response Events/Second
|
|
Starting pre-init...
|
|
The ^XUCM global will now be synchronized with VPM file numbers
|
|
Moving ^XUCM(1) to ^XUCM(8986.095)-CM SITE PARAMETERS
|
|
Moving ^XUCM(2) to ^XUCM(8986.098)-CM BERNSTEIN DATA
|
|
Moving ^XUCM(3) to ^XUCM(8986.3)-CM SITE NODENAMES
|
|
Moving ^XUCM(3.5) to ^XUCM(8986.35)-CM SITE DISKDRIVES
|
|
Moving ^XUCM(4) to ^XUCM(8986.4)-CM METRICS
|
|
Moving ^XUCM(5) to ^XUCM(8986.5)-CM DISKDRIVE RAW DATA
|
|
Moving ^XUCM(5.1) to ^XUCM(8986.51)-CM NODENAME RAW DATA
|
|
Moving ^XUCM(6) to ^XUCM(8986.6)-CM DAILY STATISTICS
|
|
This is for VMS/DSM systems only!
|
|
Sign on as your VMS SYSTEM manager and try again.
|
|
This routine will assist you in configuring TASKMAN to run from a
|
|
DCL CONTEXT.
|
|
This procedure begins on page 294 of the Kernel 7.1 SYSTEMS MANUAL.
|
|
First, select an HFS device for writing to Taskman's home directory.
|
|
Now, let's create Taskman's home directory.
|
|
SET FILE
|
|
The final step will be to define the TASKMAN batch queues for each node
|
|
in your cluster.
|
|
Enter the name of each node in your DHCP cluster. Press RETURN when finished.
|
|
Enter NODENAME
|
|
For each of the above nodes I will create batch queues.
|
|
INIT/QUE/BATCH/OWN=TASKMAN/JOB=50/ON=
|
|
TM$
|
|
START/QUE/ON=
|
|
Select a HOST FILE SERVER device:
|
|
Enter the drive/path
|
|
Enter the full path to Taskman's home directory, ie, DRIVE:[DIRECTORY]
|
|
...WRITING OUT 'LOGIN.COM'
|
|
LOGIN.COM
|
|
...WRITING OUT 'ZTMWDCL.COM'
|
|
ZTMWDCL.COM
|
|
...WRITING OUT 'ZTMSWDCL.COM'
|
|
ZTMSWDCL.COM
|
|
This step creates a new entry in UAF called TASKMAN.
|
|
You will need to provide the UIC code in the format '[#,#].'
|
|
Taskman will require at LEAST the following privileges:
|
|
CMKRNL, TMPMBX, OPER, NETMBX
|
|
Would you like to see a brief listing of UAF records
|
|
Enter a UIC code in the format '[#,*]' to restrict listing by group
|
|
MC AUTHORIZE SHOW/BRIEF
|
|
Assign TASKMAN to what UIC
|
|
Would you like to copy an existing user over to TASKMAN
|
|
Enter the Username or UIC Code to copy
|
|
MC AUTHORIZE COPY
|
|
MC AUTHORIZE ADD TASKMAN/UIC=
|
|
on node
|
|
Performance Monitor graph
|
|
Mean:
|
|
Plot raw data for which drive?
|
|
Select the NODENAME from which data was collected:
|
|
Select a metric to plot:
|
|
To sort by a particular nodename, enter that node:
|
|
Press return, or select a drive to sort on:
|
|
PLOT BY METRIC
|
|
NEED A DUZ...
|
|
This routine allows you to define the site file parameters
|
|
needed to run VPM, then instructs taskman to create the
|
|
host system directories for data files and system-specific
|
|
command procedures.
|
|
HOST FILE DIRECTORY UNDEFINED!
|
|
HFS DEVICE UNDEFINED!
|
|
CM NODENAME FILE IS INCOMPLETE!
|
|
VPM INSTALL
|
|
This final step will queue a task to install/rebuild the VPM com files.
|
|
This installation has been queued to task number:
|
|
CREATE/DIR
|
|
XUCMVPM.COM
|
|
UNABLE TO OPEN FILE
|
|
$!XUCMVPM - VA PERFORMANCE MONITOR DATA COLLECTION
|
|
$!THIS VERSION CREATED
|
|
$ SET NOON
|
|
XUCMMONITOR.COM/NOPRINT
|
|
XUCMMONITOR.COM
|
|
XUCMMONITOR.COM<POP>-
|
|
$!XUCMMONITOR.COM - MONITOR VAX METRICS
|
|
$!THIS VERSION CREATED ON
|
|
$ MONITOR DISK/ITEM=ALL,SYSTEM,STATE,LOCK,DLOCK -
|
|
MONITOR_'NODE'.DAT/NODISP/INT=10/END=
|
|
$ OPEN/APPEND MON
|
|
MONITOR_'NODE'.DAT
|
|
$ SHOW USER/INTERACTIVE/NODE='NODE'/OUT=MON
|
|
$ CLOSE MON
|
|
$ SHOW MEMORY/OUT=
|
|
MISC_'NODE'.TMP
|
|
$ OPEN/APPEND MEM
|
|
$ WRITE MEM
|
|
HARDWARE TYPE=''TYPE'
|
|
$ SHOW DEVICE D/FULL/OUT=MEM
|
|
$ CLOSE MEM
|
|
MISC*.TMP
|
|
$!XUCMRTH.COM *TOOLKIT VPM*
|
|
/MAN RTH^%ZOSV2
|
|
VPM INSTALL COMPLETED
|
|
Would you like to start over
|
|
MONITOR*.TMP.*
|
|
<VPM 3-SD WARNING>
|
|
on node
|
|
Compute
|
|
Compute (Out
|
|
ENQs Forced
|
|
DISTRIBUTED LOCK MANAGEMENT
|
|
New ENQ Rate
|
|
Block
|
|
Interrupt
|
|
Kernel
|
|
User Mode
|
|
Idle Time
|
|
Process Count
|
|
Page Fault
|
|
Page Read
|
|
Free List
|
|
Modified List
|
|
Direct I/O
|
|
Buffered I/O
|
|
number of processes
|
|
Capacity Mgmt Statistics
|
|
Your task has been queued:
|
|
ON NODE <<
|
|
Summation of Drive Activity for all Nodes
|
|
Enter EARLIEST DATE:
|
|
Enter LATEST DATE:
|
|
Mail report
|
|
XUCM*
|
|
VPM COMPACT SITE SUMMARY
|
|
Queue for what DATE/TIME?
|
|
metric not in file!
|
|
Model
|
|
VPM Summary for
|
|
DKQ
|
|
VolSet
|
|
Total MB
|
|
MB Free
|
|
VPM Compact Summary for
|
|
VPM Compact Summary
|
|
Select VPM Site
|
|
is not a VPM site.
|
|
No mailgroup defined in VPM site parameter file, no VPM daily report
|
|
Enter Ending Date
|
|
Ending Date must be Greater than Beginning date.
|
|
XUCS*
|
|
COMMANDS VS GLOBAL REFERENSES REPORT
|
|
** Summary For This Session **
|
|
Total Routine Commands Counted:
|
|
Average Routine Commands per Second:
|
|
Total Routine Global References Counted:
|
|
Average Routine Global References per Second:
|
|
CMNDs vs GREF By Routine
|
|
Note: (???) indicates that NEGATIVE counts occur in the MSM-RTHIST global.
|
|
Ave.# Jobs:
|
|
Routine CMND/S
|
|
GREF/S
|
|
# of Routines <thresh:
|
|
Tot. CMDS <thresh:
|
|
Tot. GREFs <thresh:
|
|
Tot. Grefs <thresh:
|
|
GLOBAL REFERENSES REPORT
|
|
Total Global References:
|
|
Average Global References per Second:
|
|
Global Reference Report
|
|
# of Globals <thresh:
|
|
SYSTEM STATICAL REPORT
|
|
MUMPS Commands
|
|
Global Set Commands
|
|
Global Kill Commands
|
|
Global Fetches
|
|
TOTAL Global Accesses
|
|
TOTAL Logical Read Operations
|
|
TOTAL Logical Write Operations
|
|
Disk Reads
|
|
Disk Writes
|
|
TOTAL Disk I/O
|
|
TOTAL DDP Global Requests
|
|
TOTAL RVG Requests
|
|
RVG Hits
|
|
RVG Cache eff.
|
|
INPUT From Terminals
|
|
#################### #################### ####################
|
|
#################### #################### ####################
|
|
#################### #################### ####################
|
|
#################### #################### ####################
|
|
#################### #################### ####################
|