After a long gestation ProTop 3.3s has arrived!

Highlights include (but are by no means limited to):

• Tips and Tricks on the splash screen
• A startup warning if you do not have -tablerangesize and -indexrangesize set adequately
• A new screen to show the *rangesize calculations (press "1")
• Alternate Buffer Pool indicator columns on table and index activity
• A new "2" data collector to show buffer statistics (some of which are only available in OE 11.7+)
• New "user experience" metrics - this metric reads a fixed number of records and times it
• Modified the startup handling of .pf files to increase flexibility
• Modified UIO to show the current number of locks held per session rather than lock requests/sec (11.7+)
• Generally took advantage of OE 11.7 improvements where available
• Added the option to toggle the client statement cache on/off for a session when selecting a session via # or P
• Added an experimental "task scheduler"

... and lots of other stuff!

Done.

I dunno, you're a pretty shady character...

... and I have now added that capability! Coming soon

I had a bad result with this - it seems to interfere with the icons somehow.

(Yes, that might mean that I am testing on Windows. For the record I regret it already.)

As Paul says every block with data in it must be read - even for an incremental backup. Which is why an incremental backup takes about as long to run as a full backup.

-Bp 10 helps to prevent the entire -B buffer pool getting flushed as a result of a backup running and the follow-on impact that that will have to the hit ratio. During the backup itself though you will still almost certainly see a poor hit ratio.

I'm circling back around to this today and as I poke around in the code I find that I actually started on this and had many of your examples covered back on June 7th -- three weeks before you asked

I have added an "on stop, undo leave" around most of protop.p which I think will take care of this problem.

A new release is imminent - there is really just one more thing I'd like to squeeze in.

The parenthesis are a bad idea. You really want to do it like this:

if "%DLC%"=="" set DLC=D:\Progress\OpenEdge11.7

Portal v4 has a lot of enhancements to stuff like this. Real Soon Now.

As Paul says every block with data in it must be read - even for an incremental backup. Which is why an incremental backup takes about as long to run as a full backup.

-Bp 10 helps to prevent the entire -B buffer pool getting flushed as a result of a backup running and the follow-on impact that that will have to the hit ratio. During the backup itself though you will still almost certainly see a poor hit ratio.

All of the latest ProTop news! https://wss.com/protop-newsletter-feb-2021/

@wim-van-der-ham Simple solution: time to update ProTop

FYI - that issue was fixed in ProTop 3.1415. Release notes item #44

What version of Progress? Which release of ProTop?

ProTop 3.1415 is now available for download!

Highlights include, but are not limited to:

• Moved a lot of functions to different hot keys. You may want to study the help screen ("?") before you report that something is missing.

• Improved the labelling of various fields on the dashboard and configuration screens.

• Added self-monitoring. Uncomment the logmon.sh scheduled task in etc/schedule.cfg to enable self-monitoring of tmp/*.debug files.

• Added quite a few color highlights to draw attention to interesting metrics. Yellow = warning, Red = alert, Green = Good job!.

• Added a "dbIdle" column to "Active Transactions" (the "x" command). This field shows how long it has been since the session has accessed the database. This should help to identify long running transactions that are "just sitting there" (probably on a screen waiting for user input). The dbIdle can only be measured from the point where the "x" screen was invoked - so it is possible that a session that already had an active transaction could really be idle longer than is shown. (This is similar to the blocked session duration on the "b" screen.)

• Improved the prompting for table and index names in the "users of" screens to prompt if the corresponding data has not been previously supplied when they are invoked.

• Added a file viewer to enable viewing of reports from within ProTop ("v")

• Finally figured out how to avoid the "selected" row on the browse widgets -- which means that there are now no more unhighlighted cells!

• Added a localenv variable to let users control colors. Check it out! Take a look at "0" to fiddle with colors inside the UI in order to decide what you like. (The "0" key may change somewhere down the road, but I couldn't think of a better key to attach it to right now.)

• Got to the bottom of the "error 49" problems when adding or removing panels to display. It turns out to be a Progress bug It is expected to be fixed in the next patch for 11.7 and 12.2. See NOTES, #28

• Added bin\pt3win.bat. This script will run prowin.exe rather than _progres.exe (prowin.exe does not suffer from the bug above). It is less ugly than I expected. Do not expect any fancy GUI stuff to happen. There is no mouse support and if you get all clicky bad things are likely to happen. Do NOT use this as an excuse to make client-server connections.

• Made the "dashboard", "configuration", and "osinfo" panels the same height. This makes it so that browse panels do not get resized when switching between "top of screen" viewer panels. This, in turn, makes the impact of error 49 (above) slightly less severe when using the default startup panels. Changed the command key for osInfo to lower-case "o" in order to make it consistent with "d"ashboard and "c"onfiguration.

• Re-arranged the configuration panel to better group items. Added fields for -hashlatchfactor, -lkhash -cpterm. Removed 2PC and Failover Clusters. Changed from "excess" table and index to "unmonitored".

• Messed around with osInfo. Since it got taller (see item #30) there is a lot of empty space so I added some fields. Thinking about adding more. Revisited AIX and HPUX and updated them.

• Added 24 bit RGB colors to etc/protermcap.protop. (More details in NOTES file.)

• Expanded the width of aiinfo "size" field to 1TB and added a "last op" field to show when extents have been switched, marked empty etc.

• Added a new set of alerts for "slow after-image extent switching". If your IO subsystem is slow and klunky (i.e. SAN storage) and it sometimes takes a long time for the AIMGT daemon to copy FULL extents before marking them EMPTY then this alert is for you.

• Added a "CSC Age" column on screens that show client-statement-cache information. This column tells you how old the CSC data is. (If the same statement is executed in a loop the CSC timestamp does not update - this is NOT a ProTop error.)

• The default etc/schedule.cfg will now automatically purge and cleanup TMPDIR and LOGDIR weekly at 2 minutes past midnight Sunday.

• Winnowing away as much conditional compilation code as possible. The number of &IF pre-processor statements in the code has been greatly reduced. This doesn't directly impact functionality but indirectly it means that the code is more resilient when things like "proutil updatevsts" have not been run since the usual alternative for things like &IF PROVERSION... is to switch to dynamic queries for the bits that are being handled conditionally.

Additional gory details can be found in the NOTES files.

tick, tick, tick...

The time for pre-release feedback is drawing to a close!

ProTop 3.1415 is almost ready to escape into the wild! If you would like to test the release candidate you can easily obtain by using the pt3upd.sh (or pt3upd.bat) command like this:

tom@protoptest> pt3upd.sh pt3.1415x
Wed Apr 29 13:24:00 UTC 2020 Updating to pt3.1415x
Wed Apr 29 13:24:00 UTC 2020 pt3.1415x-src.tar
Wed Apr 29 13:24:00 UTC 2020 curl -L -o /home/tom/tmp/pt3.1415x-src.tar.gz http://demo.wss.com/pt3upd/pt3.1415x-src.tar.gz
29/04/2020 13:24:05.655+00:00 getfile: demo.wss.com|/pt3upd/pt3.1415x-src.tar.gz|/home/tom/tmp/pt3.1415x-src.tar.gz
29/04/2020 13:24:05.655+00:00 getURL(): newSocket( demo.wss.com 80 ) size = 118
29/04/2020 13:24:05.657+00:00 connect( -H demo.wss.com -S 80 )
29/04/2020 13:24:05.661+00:00 getURL(): newSocket( demo.wss.com 80 yes )
29/04/2020 13:24:07.171+00:00 HTTP/1.1 200 OK
pt3inst.sh
protop.p
...
ustub/zipenv.x
ustub/zippyng.sh
ustub/portalenv
ustub/pt3inst.sh
Wed Apr 29 13:24:12 UTC 2020 running silent install, detailed log is in pt3inst.log
tom@protoptest> 

The release notes can be found in etc/NOTES_3.1415.

Unless someone finds a really horrible bug this will probably be made available for general release next week!

Are you suggesting that something like:

> protop

Please specify a friendly name in etc/dblist.cfg or a full database pathname.
For example:

    protop s2k

or

    protop /db/sports2000

Would be a better error message?

if ( not _Connect-device matches "*batch*" ) and _connect-batch <> "yes" then tt_Dashboard.con_self = tt_Dashboard.con_self + 1.

So "self" is counted for non-batch self-service connections. It (mostly) equates to local interactive connections.

Somehow I suspect that you are going to suggest that that might be sub-optimal