home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
OS/2 Shareware BBS: 8 Other
/
08-Other.zip
/
watchit2.zip
/
MANUAL.INF
(
.txt
)
< prev
next >
Wrap
OS/2 Help File
|
1994-07-14
|
75KB
|
465 lines
ΓòÉΓòÉΓòÉ 1. Introduction ΓòÉΓòÉΓòÉ
WATCHIT
"Make your server serve"
WATCHIT is a collection of programs to monitor LAN Server performance and
capacity. Server operational statistics are collected for analysis. You will
find a brief discription of each program in the following frames. You may
reproduce and give this demo to anyone else who shows an interest. You may not
reverse engineer this demo or use it in a productive or profitable manner.
We have included an analysis program and sample data for you to TAKE A TEST
DRIVE !
ΓòÉΓòÉΓòÉ 2. COPYRIGHT and Acknowledgments ΓòÉΓòÉΓòÉ
Copyright (C) 1993,1994
Client Server Networking Inc
P.O. Box 370111
West Hartford Ct 06137-0111
(203) 233-2951
Acknowledgements and trademarks
IBM, OS/2, and LAN Server are registered trademarks of International Business
Machines.
All other trademarks are the property of their respective owners and are hereby
acknowledged.
ΓòÉΓòÉΓòÉ 3. What is WATCHIT ΓòÉΓòÉΓòÉ
WATCHIT
"Make your server serve"
Tune IBM LAN SERVER. "Make your server serve" Collect statistics about resource
and user activity. Easily analyze them and improve performance and anticipate
capacity problems.
WATCHIT automates collection of IBM LAN SERVER capacity and performance
statistics. WATCHIT will help you quickly and easily improve performance and
anticipate capacity problems. Collection may be automated at the server or
manually controlled from your desktop. Thresholds will generate warning
messages during data collection or notification via generic alerts, NET.ERR
logging or simple messages. WATCHIT graphs mean response, sessions, shares,
bytes transferred, print jobs, big buffer and request buffer allocations.
Graphs direct you to the detail information needed to improve performance and
expand capacity. IBMLAN.INI parameters can be compared to the collected
statistics to assist you in making the proper settings. Session analysis
provides information to determine the number of connections required and file
opens. Analyze ALIAS access. Statistical data is exportable for use in your
favorite spreadsheet or database for further analysis or archiving. The
on-line manual includes operation suggestions to help you "make your server
serve".
ΓòÉΓòÉΓòÉ 4. Version 2 enhancements ΓòÉΓòÉΓòÉ
Version 2 enhancements include:
1. Data collection performance is substantially improved.
2. Remote and local data collection no longer have content restrictions.
3. Collection data for the same time period is 1/3 to 1/2 the space of version
1.
4. The number of setup programs is reduced providing for easier operation.
5. ALIAS analysis has been expanded.
6. Session analysis has been expanded.
7. Data collection for remote or local is done with the same program.
8. Drag and drop support makes analysis quicker
9. Summary information of the analysis programs can now be exported to your
favorite spreadsheet or database.
10. A remote install program is provided to assist in accessing WATCHIT on your
desktop across the network.
11. Excessive activity can now be recorded to the error log or forwarded to
NETVIEW.
12. LAN SERVER 3.0 and 4.0 are supported
ΓòÉΓòÉΓòÉ 5. Beta Tester's comments ΓòÉΓòÉΓòÉ
WATCHIT
"Make your server serve"
"A must for the LAN Server administrator.... it has helped me pinpoint problems
with my LAN that other products I've used were not aware of."
"We want to offer central management for departmental LAN SERVERS. WATCHIT can
provide the insight into local operations to make that feasible."
"Can't beat this product! It does more then others costing thousands more do
plus some !"
"We manage 48 remote LANs from a central site. WATCHIT will assist us to
continue to provide consistent LAN SERVER service without remote
administrators."
"The ability to store performance data in an archive file that can be evaluated
at my leisure is a big plus for me"
ΓòÉΓòÉΓòÉ 6. To reach us for questions: ΓòÉΓòÉΓòÉ
Client Server Networking Inc
P.O. Box 370111
West Hartford Ct 06137-0111
(203) 233-2951
IBM TALKLINK ID DEV2401
.. or WATCHIT Forum
9:00Am to 5:00 Pm Eastern time
Ask about a site or enterprise licence
ΓòÉΓòÉΓòÉ 7. Ordering Information: ΓòÉΓòÉΓòÉ
Use the SERVICES option on the menu bar to
Copy to a file or print order form.
To CLIENT SERVER NETWORKING
P.O. BOX 370111
West Hartford, Ct 06137-0111
(203) 233-2951 Visa or MC accepted
FROM:
Name:
__________________________________________
Company:
__________________________________________
Street:
__________________________________________
City, State
_________________________________________
Postal/Zip Code: __________
Telephone: (_____) ___________________
Credit card Card# __________________
Exp Date _/__/__
FEES: Per server with any number of clients
Introduction until 1/1/95
$249 X ___ Number of copies = _______
WATCHIT V2 after 1/2/95
$349 X ___ Number of copies = _______
WATCHIT Upgrade (send copy of diskette label)
$149 X ___ Number of copies = _______
IF Shipped to Connecticut add 6% Sales TAX
Total ==========
Prices include Shipping.
ΓòÉΓòÉΓòÉ 8. Installation Requirements: ΓòÉΓòÉΓòÉ
There are two installation programs provided. One is for your server and the
other for the desktops of your administrators.
Server installation requires:
1. OS/2 2.0 plus
2. IBM LAN SERVER 3.0 or 4.0 ( has been used on earlier releases )
3. Initial space requirements are at least 1.5 megabytes.
Remote:
WATCHIT access can only be installed on your OS/2 2.0 desktop after the server
installation is completed. A folder containing the program objects and data
will be made available to access the server. This remote access can be
installed on any number of administrators desktops.
ΓòÉΓòÉΓòÉ 9. WATCHIT package ΓòÉΓòÉΓòÉ
WATCHIT
"Make your server serve"
WATCHIT consists of one data collection program and four analysis programs.
ΓòÉΓòÉΓòÉ 9.1. WATCHIT (Data collection) ΓòÉΓòÉΓòÉ
WATCHIT may be operated on the server or remotely on any OS/2 requester that is
logged on with full administrator privileges. WATCHIT is the data collection
and monitoring program of the application. It provides two methods of
performing this function. Manual collection allows you to start and stop data
collection and monitoring at your discretion. Batch collection provides
automatic collection via the 'AT' command.
As the statistics are collected you will be able to observe the following LAN
SERVER activity.
Average response (ms)
Bytes sent and received by the server (k)
Date and time on the server
Number of active sessions
List by name of all logged on users.
WATCHIT Configure
Statistic collection can be configured to meet your requirements.
You may select maximums for notification and control the number of collection
cycles and duration.
Thresholds can be set to dynamically notify you when the server exceeds:
System Errors - LAN Server reported system errors.
Byte Traffic - Total bytes sent and received exceed this value.
Sessions - When the number of sessions reaches or exceeds this value.
Mean Response - Excessive mean or average response time.
Request Buffers - Failed to allocations.
Big Buffers - Failed to allocations.
Collection Notification
When the thresholds are exceeded you may select the method of notification:
Local Popup - Local popup windows.
Log to NET.ERR - Provide an entry in the local NET.ERR log.
Message Alert - Generate a Message Alert.
Generic Alert - Generate Generic Alert NET3101E.
ΓòÉΓòÉΓòÉ 9.2. ANALYZE ΓòÉΓòÉΓòÉ
ANALYZE is the primary program of understanding the data you collected. This
program has been supplied with sample data for you to "test drive".
Analyze provides a display of the operational configuration that was in effect.
These values are broken into two components: capacity and performance.
You may chart multiple measurements and usage parameters. Each chart is
selected separately and you may display as many as you need.
Response: Mean Response indicates how long each request takes to service.
Sessions: Sessions are the number of active sessions that the server has
outstanding.
Shares: Shares indicate the number of outstanding shares the server has.
Traffic: Byte traffic is the total bytes transferred during the sample.
Print Jobs: Print jobs is a count of print data being spooled to the server.
Req Buffer Allocations: Request buffer allocations is the number of times the
system has run out of REQ BUFFERS.
Big Buffer Allocation: Big buffer allocations are the number of times the
system has run out of BIG BUFFERS.
When you locate a potential problem you may select the detail related to that
sample.
Your analysis may be extended by exporting a summary of the samples to your
favorite spread sheet or data base.
ΓòÉΓòÉΓòÉ 9.2.1. Using the demo data: ΓòÉΓòÉΓòÉ
DATA ANALYSIS:
1. Using the File menu Open SAMPLE.DAT.
2. From the Detail menu select Configuration. There are two scrollable list
boxes, one for capacity and the other performance. They are the IBMLAN.INI
parameters in effect for the sample.
3. From the Detail menu select Chart Response - Notice sample 10 has the worse
response of the day.
4. From the Detail menu Chart Traffic - it was not a traffic problem.
5. From the Detail menu Chart Sessions - nothing unusual.
6. Scroll through the Detail list box on the main window to sample 10 and
double click - There were three System errors! If you issue a NET ERROR in
this sample you would have find out what happen!
7. Look at the Response chart again - we suggested samples 10 and 11 are over
the standard deviation.
8. Scroll through the Detail list box to sample 11 and double click - There
was another system error! Since the error persisted over 4 minutes check it
out! Issue a NET ERROR command and match up the times.
ΓòÉΓòÉΓòÉ 9.3. ALIAS analysis ΓòÉΓòÉΓòÉ
ALIAS provides analysis of ALIAS usage.
ALIAS - Your defined name. Only active ALIASs will be displayed.
PATH - Fully qualified path of the ALIAS.
COMMENT - The comment entered when defining the ALIAS.
MAX USERS - The number of users defined to access this ALIAS.
CURRENT USERS - The number of users who are currently referencing this ALIAS.
ALIAS report
When you end ALIAS you may save a summary report of the ALIAS activity for
futher analysis.
***** SHARE SUMMARY ****
** collection parms **
Date 02/24/94 Time started 10:14:52
Frequency every 120 seconds
Collection was manual
Server = PRCSDS02
ALIAS PATH Comment Max Users 1 2 3
$APPS D:\SERVER\APPS Applications 65535 13 13 13
$BCPPOS2 F:\OS2DEV\BCPPOS2 BorlandC 25 4 4
$BOOKMGR E:\BOOKMGR IBMBookManager 65535 12 12 12
$TIPSTST D:\TIPS\TEST TIPSTesting 65535 2 2
$VIRSCAN D:\SERVER\VIRSCAN OS/2VirusScan 65535 1 1
**** Sample Cross Reference ****
Time Mean Time File Access KBytes Sent Kbytes Recv
1 05:13:24 4 124 3521313 719589
2 05:14:45 3 143 6445177 693793
3 05:16:45 3 31 3778431 1721439
ΓòÉΓòÉΓòÉ 9.4. SESSION analysis ΓòÉΓòÉΓòÉ
SESSION provides analysis of user activity to understand current resource
usage.
User ID - Logged on ID of the user.
Machine - The name of the machine the user logged on from.
Release - The release level of the requester.
Connects - The number connections this user is holding.
Opens - The number of files open at the time the sample was taken.
SESSION Report
When you end SESSION you may save a summary report of LOGON activity.
***** LOGON SUMMARY ****
** collection parms **
Date 02/24/94 Time started 10:14:52
Frequency every 120 seconds
Collection was manual
Server = PRCSDS02
CONNECTIONS / FILES OPEN
COMPUTER UserID Session Type 1 2 3
HDAVIS1 HDAVIS OS/2 LS 3.0 1 1 1
0 0 0
JOHNMAN JMANNING OS/2 LS 3.0 1 1 1
0 0 0
S7354573 MWILLCOX DOS LS 1.2 1 1 1
0 0 0
**** Sample Cross Reference ****
Time Mean Time File Access KBytes Sent Kbytes Recv
1 05:13:24 4 124 3521313 719589
2 05:14:45 3 143 6445177 693793
3 05:16:45 3 31 3778431 1721439
ΓòÉΓòÉΓòÉ 9.5. INI PEEK ΓòÉΓòÉΓòÉ
INI PEEK provides a comparison of some of your IBMLAN.INI entries and the
server operation under load.
MAXCONNECTIONS - The number of NET USES the server can handle.
MAXOPENS - The number of files that can be open on the server.
MAXSHARES - The number of resources the server can share.
MAXUSERS - The maximum number of users who can access the server.
NUMBIGBUF - The number of 64k buffers used for large file transfers.
NUMREQBUF - The number of 4k buffers used to handle requests.
ΓòÉΓòÉΓòÉ 10. Tune OS/2 information ΓòÉΓòÉΓòÉ
WATCHIT provides additional information to help you tune OS/2. Subjects
discussed are:
Bottle necks
MEMORY
DISK
CONTROLLER
DRIVE
PLACEMENT
CACHE
CPU
OS/2 BASICS
SWAP
FAT or HPFS
CONFIG.SYS tune up
ΓòÉΓòÉΓòÉ 11. Tune LAN SERVER information ΓòÉΓòÉΓòÉ
WATCHIT provides additional information to help you tune LAN SERVER. Subjects
discussed are:
Entry vs. Advanced Server
PERFORMANCE
AUDITING
ALERTSCHED
MAXSESSREQS
NUMBIGBUF
NUMFILETASKS
NUMREQBUF
SIZEREQBUF
SRVANNDELTA
SRVANNOUNCE
SRVHERISTIC
CAPACITY
AUTODISCONNECT
CLEANUP
MAXCONNECTIONS
MAXOPENS
MAXLOCKS
MAXSEARCHES
MAXSESSOPENS
MAXSHARES
MAXUSERS
SESSIONS
LOGON speedup
Large Domain Tuning Update
ΓòÉΓòÉΓòÉ 12. Other products: ΓòÉΓòÉΓòÉ
CONDUIT is a no nonsense LAN file distribution system that facilitates
administration, distribution, and inventory management for OS/2 2.0 and DOS
administrators and clients.
CONDUIT reviewed as "elegant in its simplicity" can be summoned from the Work
Place Shell as a powerful application management tool.
CONDUIT automates tedious and time consuming desktop management. Without a
mainframe or mini CONDUIT can do installation, upgrades, version control,
license supervision on any number of networked client personal computers. It
operates with IBM, Novell, Banyan, Lantastic, and Microsoft products using a
unique network independent design.
CONDUIT defines and controls distribution. It provides built-in security
safeguards with identification and passwords; creates directories; remotely
executes programs; develops client profiles; and maintains inventory data.
CONDUIT defines and controls distribution. It provides built-in security
safeguards with identification and passwords; creates directories; remotely
executes programs; develops client profiles; and maintains inventory data.
Collected profile data will allow you to reconstruct CONFIG.SYS, AUTOEXEC.BAT
or STARTUP.CMD files that have become corrupted.
Distribution tracking allows consolidated or separate LOGs. The log tracks
accesses, distribution status and error conditions such as missing files, ID,
Password or space.
All of CONDUIT'S features are available today for only $299 per server
regardless of the number of clients with no "pak" or limit restrictions.
CONDUIT is the easiest to use state-of-the-art LAN file system and is the best
price performer on the market today and in the future.