Release Notes 3.3s

-1) In case you've skipped a few releases -- ProTop ChUI now requires a
minimum screen size of 160 columns by 48 rows -- 4x an old 80x24 green

0) Don't get used to this. Really. It is just a coincidence that these
release notes are getting created. You have been warned.

1) Cleaned up the server name detection:

  a) Check the PTSERVER environment variable (set in bin/localenv). This
  is helpful if you want to use a "cluster name" rather than a "node name"
  in a clustered environment.

  b) If no PTSERVER is specified use the 2nd field in "uname -a" (Unix) or
  the "hostname" command on Windows.

  c) If etc/protop.cfg contains an entry for "server" use that (overrides
  all of the above - not terribly flexible).

The server name is used when deciding which etc/dblist.cfg lines are
relevant for an instance of dbmonitor.

2) Added alternate buffer pool information to "Area Info", "Table
Activity", and "Index Activity".

3) Added a new data-collector for "Buffer Pools" that can be triggered
via the "2" key.

4) Added seqMaxPct to monitor how close a sequence is coming to its
maximum value and seqPct32 to identify sequences approaching the old
32 bit limit -- which can be an issue if old code is updating INTEGER
fields rather than INT64. 

5) Added user defined metrics to the dashboard.  See lib/usermonx.p
for details.

6) Fixed a bug in the calculation of pctLastX (the percentage full of
the last fixed extent).

7) Expanded the width of the "Q" (SQL users) screen.

8) Expanded decimal places in percentage fields of Big B Guesstimator.

9) Tweaked some unplanned features in "users of table" that might have
sometimes possibly maybe resulted in potentially confusing data being
occasionally shown under some circumstances.

10) added a "user experience" metric (aka "zippy") to time access to a
standard set of db records.  Think of this as "records per second" as
seen by a random user connecting via shared memory.

11) HTML encoded the email Id provided during installs to support people
who like to embed crazy stuff like "+" in their email addresses.

12) Added "idx reads" and "idx deletes" to the dashboard screen.

13) Added ".lg file age" to configuration to show how old the oldest
message in the .lg file is.

14) Jumped through hoops to support 32 bit _progres executables if you
happen to be running 32 bit db servers on Windows post 11.3. Which is
a ridiculous thing to be doing if anyone wants my opinion...

15) Added optional .pf file support at pt3agent startup for:


if these .pf files exist they will be added to the pt3agent startup
command in the order shown.  The db name should NOT be included in
these parameter files as it is independently added to the startup
parameters.  Take note: etc/ is NO LONGER SAFE for your
localizations (such as -U and -P for example).

16) Improved handling of startup errors. Now if something goes horribly
wrong during startup ProTop should just exit rather than getting hung
in a loop continually trying to start.

17) Added the "ProTop Name" to _connect-usermisc for ProTop session.
This makes it easier to see if the protop.p, pt3agent.p or other
ProTop programs are having an impact.  (oe11.7+)

18) Added "P" to client type codes to indicate PASOE agents.

19) Added support for new 11.7+ _actBuffer data. The "2" screen now has
extended information on the types of blocks in buffer pools:

        Label     Meaning
        =====     ==============
        Pool      Which buffer pool?  Primary or Alternate?
        Tbls      How many tables are *assigned* to this pool?  (even if no -B2 is active they are still *assigned*)
        Idxs      How many indexes are *assigned* to this pool?  (even if no -B2 is active they are still *assigned*)
        LOB       How many LOBs are *assigned* to this pool?  (even if no -B2 is active they are still *assigned*)
        LRU Stat  What is the status of the LRU for this area?
        Buffers   How many -B or -B2 buffers have been allocated?
        Unused    How many allocated buffers are unused?
        Active    How many allocated buffers are being used?
        DataBlk   How many buffers are data blocks?
        IdxBlk    How many buffers are index blocks?
        Mstr      How many buffers are master blocks?
        SEQ       How many buffers are sequence blocks?
        Area      How many buffers are area blocks?
        Ctrl      How many buffers are control blocks?
        Obj       How many buffers are object blocks?
        OLst      How many buffers are object list blocks?
        OCL       How many buffers are object control list blocks?
        CMap      How many buffers are cluster map blocks?
        BMap      How many buffers are block map blocks?
        logRd     How many logical reads per second?
        OSRd      How many os (disk) reads per second?
        logWr     How many logical writes per second?
        OSWr      How many os (disk) writes per second?
        Hit%      What is the hit ratio?

20) Replaced call to "rfutil" with new 11.7+ _areaStatus info for AI

21) Modified checkpoint screen to reflect flexible number of checkpoints
and new fields added with 11.7+.

22) For 11.7 and higher the "Rec Lk" column on the UIO and SQL screens
shows the current number of locks. (Prior to 11.7 it shows the number of
lock requests per second.) This should help considerably in monitoring
users who obtain large numbers of locks. "recWts" has been replaced, in all
releases, with "lkHWM" - the High Water Mark of lock entries used by a
session.  (Requests per second is still useful in earlier versions.)

23) ProTop will now over-write the "standard" files in etc. It is no longer
safe to put localizations into files such as etc/ or pt3agent.cfg
and so forth. See #15 for .pf file, for .cfg files follow the established
convention of pt3agent.XXX.cfg where XXX is the custId, serverName or

24) More revisions to the the "excess" range calculation.  The "1" command
will now bring up a screen with detailed information about how the excess
is calculated along with some suggested values for setting -basetable,
-baseindex, -tablerangesize and -indexrangesize.

25) Simplified temp-table debugging and data collection examples by moving
the setting of DEBUGTT to lib/ttdebug.i, in this way you only need to
modify a single include file to enable the sample code.

26) Revised configuration screen to use _dbParams when it is available and
to avoid screen scraping PROMON or parsing the .lg file if possible.

27) Added "message of the day" helpful hints, useful tips and similar
stuff at startup.  The marketing droids may eventually get wind of this
and start dropping hints.

28) Added "inodes" to the df screen.  Currently only populated for Linux
and AIX (AIX is untested).

29) Added an experimental "task scheduler".  See etc/schedule.cfg for
details.  It seems to work on Linux.  Windows testing has been "light"
(IOW -- "it compiles")..

30) Modified Windows protopenv.bat to only use the "install value" of
DLC if the DLC environment variable has not already been set.  This is
intended better to support situations where there are multiple versions
of Progress installed on Windows.

31) Added an option to turn on client statement caching when selecting a
user via either "#" (by usr#) or "P" (by PID). This is a lot less invasive
than turning it on for *everyone* and a lot easier than going into PROMON
to focus on a small number of users.

32) Added "TRX Lag" to the replication agent screen.  This shows the time
difference since the last acknowledged replcation block.  If there are no
active transactions TRX Lag will be 0.

33) ChUI ProTop will warn on startup if the -*rangesize parameters are