Hello Collectors,
The newest version of collect v1.11, is now available (see below
for locations). This will be the last version from me, however,
it seems likely that collect will be taken over by the Performance
Manager group, led by Tina Anderson. Please contact her at
tea_at_zso.dec.com with questions, comments, and recommendations.
NOTES:
[steel]
There is still no support for LSM or RAID (SWXCR) disks.
I suspect that there may be minor problems on BL21 with the
tape subsystem. If you have problems, contact me to find out
how to patch the sources (it's minor).
[V4.0x]
collect will probably need to be modified to take advantage of
the patches which will (finally!) fix the unreliable disk
statistics bugs. (%BSY > 100 and so on).
take care,
Rob Urban
[the README for V1.1 follows]
##########################################################
## PLEASE READ "Disclaimer" BEFORE USING THIS SOFTWARE. ##
##########################################################
========================
= This is version 1.11 =
========================
[I no longer work for Digital, so I will not be supporting collect anymore.
please contact Tina Anderson (tea_at_zso.dec.com) with questions, answers, bug
reports, and requests for functionality]
DISCLAIMER
==========
This is TOTALLY UNSUPPORTED by Digital. I have done my best to make it
useful and accurate, but it is in no way guaranteed to do anything, except
take on some space on your hard disk :-). On the other hand, if you find
a bug, or think something is incorrect, send me a mail, and, time permitting,
I'll have a look. Please see Doc/Disclaimer.
STEEL (Dunix V5) Notes:
=======================
LSM is not supported on steel.
RAID disks (SWXCR) are not supported on steel.
INTRODUCTION
============
This is a data collector for Digital UNIX V3.x and V4 for operating system
and process data. See the manual page 'collect(1)' for more info.
KIT LOCATION
============
It can be retrieved as follows:
internally: nasaxp.rto.dec.com:collect/
externally: ftp.digital.com:/pub/DEC/collect/ (US)
ftp.digital.de:/pub/DEC/collect/ (Germany)
ftp.digital.com.au/pub/unix/tools/collect (Australia)
The kits are:
COLLECT111SETLD.tar (setld kit)
collect-1.11.tar.Z (sources)
and get one of the following if you want to use the GUI or filter scripts
PERL5003SETLD.tar (Dunix V3+)
PERL5004SETLD.tar (Dunix V4+)
BINARY FILE FORMAT
==================
In collect v1.11 the binary file format has changed again (now V12), however
the older binary file format versions (>= V7) are transparently read
by collect V1.11. You should not even notice that a binary file is older.
PERL
====
The above setld-kit for perl includes the Tk extension. If you want to use
the Graphical User Interface for collect, "collgui", you MUST get this kit
and install it using the standard software installation utility "setld".
If you already have perl, you probably don't have Tk (I mean the Tk
extension for perl NOT tcl/tk). You can still use 'cfilt'. You just
need to edit /usr/bin/cfilt and set the path of perl to where you have
perl installed. If you also have Tk for perl, you need to edit
/usr/bin/collgui and set the path in the same way.
COMPILING
=========
Go the the top-level directory where you unpacked the kit and type
'make'. Assuming everything goes well, you should have a binary in
the 'src/' directory called 'collectXYZ', where XYX is the operating
system ID, such as '40B' for V4.0B, '50' for V5.0, etc.
INSTALLATION
============
If you want to install everything under /usr/local, that is
collectXYZ,scripts -> /usr/local/bin
script libraries -> /usr/local/lib/collgui
manpages -> /usr/local/man/...
then typing, as root, 'make install' at the top-level should do the trick.
Keep in mind that the executable 'collect' will be installed under
/usr/local/bin. If need collect for different versions of Digital UNIX,
you should probably rename 'collect' to something version-specific, such
as 'collect40D' (in the case of Dunix V4.0D) so that multiple versions
can be installed under /usr/local/bin/.
If you prefer to carry out the installation by hand, the following is a list
of steps you need to execute:
Collect
- -------
To collect data, you need only the collect executable that you compiled
above. You need put it in your path, say in /usr/local/bin/, and to make
it SetUID-root:
[as root]
cp src/collect(XYZ) /usr/local/bin
chmod 4755 /usr/local/bin/collect(XYZ)
chown root /usr/local/bin/collect(XYZ)
You may also want to name it simply 'collect'. That's up to you.
GUI
- ---
If you also want to be able to use the GUI, you will need perl with the
Tk extension. If you aren't sure how to do this, get the perl kit above
and install it.
Then you need to copy the relevant scripts to directories in your (and
others?) path:
[you may need to do this as root]
cp src/Scripts/{cfilt,collgui} /usr/local/bin
mkdir /usr/local/lib/collgui
cp src/Scripts/Collgui-lib /usr/local/lib/collgui
You will probably have to edit /usr/local/bin/collgui and /usr/local/bin/cfilt
and fix a few things:
for collgui and cfilt: the very first line must have the proper path to
perl, for example "#!/usr/bin/perl" in the case of my perl kit.
for collgui: the line with "use lib qw(/some/path)" must be changed
such that /some/path is "/usr/local/lib/collgui" (or wherever you put the
Collgui-lib stuff).
Manpages
- --------
Copy the manpages:
[you may need to do this as root]
cp src/Doc/*.1 /usr/local/man/man1
/etc/magic
- ----------
If you want to be able to use the 'file' command to identify binary
collect files, as root, append the contents of 'magic.rules' in the
top-level directory to your /etc/magic files (as 'root').
Done!
PERFORMANCE CONSIDERATIONS
==========================
As far as performance is concerned, the collector will only have
significant impact if you collect everything every second (i.e.:
collect[34] -i1 -fcollect.out). If you leave the process subsystem
away, or give it a different interval (-i1:10), the collector should
not consume a significant portion of the cpu.
run 'collect -h' for more information about the various switches, or look
at the manual pages, which are available for 'collect', 'collgui', and
'cfilt'.
CFILT INTRODUCTION
==================
'cfilt' is a perl script which can extract ANY arbitrary data from the
output of collect. Look at 'cfilt -h', or the manpage for more info.
Here an example (the '\' characters represent continued lines):
collect -i1 -fcoll.bin # collect into 'coll.bin'
...wait a bit :-) ...
^C # terminate
collect -p coll.bin |\ # read binary file (output ascii)
cfilt cpu:idle mem:free\ # grab cpu idle and free memory
-a5 > coll.data # average over 5 seconds
of course, I could have just specified an interval of 5 seconds in the
original collection stage, then I would not have needed to use '-a5',
but who knows what one might need? I think it's better to collect more
than I need just in case...
If you use collect, and like it, or don't like it, or have comments about
it, or want to suggest improvements... send me a mail.
Rob Urban (urban_at_spielwiese.de)
Received on Fri Jan 08 1999 - 20:37:44 NZDT