CentOS 6.7 Ksnapshot and X klunky

First, a couple minor corrections.

I erroneously provided CPU info from one of my other boxes in the OP. Actual CPU is
$ cat /proc/cpuinfo
processor	: 0
vendor_id	: AuthenticAMD
cpu family	: 16
model		: 10
model name	: AMD Phenom(tm) II X6 1035T Processor
stepping	: 0
cpu MHz		: 800.000
cache size	: 512 KB
...
cpu cores	: 6
...

I also said "openoffice" when it's actually "libreoffice".

Summarizing the original issue:
Top shows
  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
32160 root      20   0  259m 104m  22m R 76.5  1.3 559:07.79 Xorg
23391 hardtolo  20   0 6293m 215m  26m S 16.9  2.7  21:49.86 java

And several instances of Xorg (one per user in addition to the shown
root instance), java, FF, soffice.bin, a plugin-container for FF, ...

Using stock Gnome Desktop with three active X sessions, 6 desktops for
two users, multiple FF instances with multiple tabs.

... when using Ksnapshot to select a region of the
screen the "window" no longer keeps up with my mouse movements, but lags
and is "jerky"

Per Gordons's question "what's process ..." we got:
  Looks like its java plugins to run trade screens application from
  ADVFN (Investorshub). But that's no different than what I was running
  before the update to CentOS 6.7.

Gordon also asked:
  and does Xorg stop using a lot of CPU time
  if you terminate it?

I noted it should ... and I mentioned how FF, plugins-container, various web page features, etc. seemed to make FF and plugins CPU hogs. I ended with "... all this same stuff was going on before the 6.7
update. I would keep an eye on FF CPU and when it got unreasonably high
I would close pages or even end FF and restart and problems were
ameliorated for a while. ... But even in those cases I didn't see a "jerky" and lagging selection window when I was selecting a region with ksnapshot. The usual symptom was a slow redraw of FF windows ... I am *not* seeing those symptoms  with the current issue ...

I then described a step I took to see if it helped:
  Ended the trade screens applet and tried selection of a region with
  ksnapshot - same results were a lag and jerkiness. I checked all the
  open FF windows to see if I had other known hog applets running but
  didn't see any.

  Trying shutting down FF and restarting, ... region selection still
  lagging and jerky. Top shows
  PID USER     PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
32160 root     20   0  221m  75m  13m R 68.7  1.0   1192:08 Xorg
 2025 hardtolo 20   0 1479m 281m  78m R 14.6  3.6 226:35.72 soffice.bin
 1035 wild-bil 20   0 1116m 315m  43m S  7.6  4.0 196:15.91 firefox

  Xorg dropped from 76.5% at the time I posted yesterday

So that test provided minimal improvement, re CPU usage, and none re Ksnapshat region selection performance.

In a later post I noted ending FF for this user reduced CPU a few more points but left the Ksnapshot region selection unimproved.

Moving on to latest steps and results ...

In the below, sometimes the "guilty user" Xorg did not appear in the screen capture so I posted the CPU usage of "the innocent", which did appear in "top" and marked it with its PID, 12947, figuring that meant the "guilty user" Xorg was below that result. The "guilty user" PID is 12109, also marked.

Went about my normal business Friday using trades screen java applets in the browser, working spreadsheets, visiting web sites and posting on boards ... and took snapshots of "top" at arbitrary intervals.

Xorg CPU usage was 12109 40%-44% in the two snaps at 9:45 and 13:10. At 14:07 it was 12109 67.5%. Java showed ~11% to ~15% and FF showed 2%, then 10.9% then 45.3% for those snapshots. Soffice.bin showed 8.6%, 11.9% and 11.6%.

Figured that was enough baseline ...

At 14:11 (just 4 minutes later) I closed all those spreadsheets (can't remember what made me suspect them) and top showed CPU usage of 12109 Xorg 0.7%, java 10.9%, FF 1%.

This was a *huge* improvement from ending what seemed to be unrelated applications, other than whatever load would be placed up Xorg. The Ksnapshot region selection seemed a bit better, but still not as fast or smooth as prior to the 6.7 update. Note perception is everything in these last two conclusions - they are not objective measurements,

Two subsequent snapshots of "top", 15:15 and 16:07 confirmed improvement with CPU usage of: 12109 Xorg 0.3%, 0.7%; java 8.9%, 15.2%, normal ranges unchanged; soffice.bin not started in first snap, 2.7% in second.

So the next step was to try for repeatability.

Saturday's testing was unable to duplicate a couple aspects: no data was being fed to the java trades applets running in FF; much reduced browsing and interacting on message boards via FF by me.

During the following, FF ran no more than 5%, varied from 1%-5%, and plugin container <=0.3%. Spreadsheets were initially opened but no work was done in them.

Started two trades screens applets in FF and, opened one libreoffice spreadsheet CPU usage showed: 12109 Xorg 15.2%; 2 instances of java 11.2%, 10.9%, soffice.bin 3.3%.

Waited ~2 hours, opened a second spreadsheet and checked again, CPU usage: 12109 Xorg 14.2%, 3 instances of java 8.9%, 7.9%, 6.6%, soffice.bin 3%.

Waited 25 minutes and with the two spreadsheets open, CPU usage: 12109 Xorg 15.8%; java 16.8% 9.9% 8.6%; soffice.bin 4%.

Waited five minutes and opened a third spreadsheet. CPU usage: 12947 Xorg 0.3%; 3 java 12.2%, 9.9%, 8.6%; soffice.bin not seen (out of window?).

Waited 5 minutes and opened a fourth spreadsheet, CPU usage: 12109 Xorg 0.7%; 3 java 8.9%, 8.3%, 7.3%; soffice.bin not seen (out of window?).

Finished for the night, left everything as was. Next morning ...

At ~5 A.M. checked to see if leaving idle overnight had an effect (which I had considered), CPU usage: 12947 Xorg 0.7%; 3 java 11.9%, 7.3%, 6.6%, soffice.bin not seen (out of window?).

~15 minutes later, closed and reopened one spreadsheet, CPU usage: 12109 Xorg 0.7%; 3 java 12.2%, 7.9%, 6.6%; soffice.bin 0.3%.

~25 minutes later opened another spreadsheet, giving five now, CPU usage: 12947 Xorg 0.7%; 3 java 6.6%, 6.3%, 5%; soffice.bin not seen (out of window?).

Then I started working in the spreadsheets and after ~7 minutes, CPU usage: 12109 Xorg 53.8%; 3 java 10.6%, 9.3%, 7.6%; soffice.bin 11.9%.

I stopped and just let that X session go idle for about an hour, just watching it via top in another X session.

A 6:09 snapshot showed CPU usage: 12109 Xorg 54%; 3 java 11.9%, 11.3%, 7%; soffice.bin 11.3%.

~15 minutes later I opened the sixth spreadsheet, normally how many I run, CPU usage: 12109 Xorg 50.6%; 3 java 10.3%, 10.3%, 8.9%; soffice.bin 10.6%.

Let things set for about 2.25 hours, came back and I began doing a little FF stuff, working the spreadsheets, and about 40 minutes later (09:07) checked again, CPU usage: 12109 Xorg 56.9%, 3 java 10.9%, 8.3%, 7.9%; soffice.bin 13.2%.

45 minutes later, CPU usage: 12109 Xorg 58.6%; 3 java 10.6%, 9.3%, 6.3%; soffice.bin 13.2%.

Quit working the spreadsheets and FF and checked about 1.5 hours after the prior check, CPU usage: 12109 Xorg 59.3%; 3 java 10.6, 8.3% 8.3%; soffice.bin 12.6%.

Quit for the day and the next morning, after letting everything sit overnight ...

Doing a little FF stuff this A.M. FF finally got over 5% - made 6%. Plugin-container not in sight, so must be low right now.

5:43, after sitting idle overnight, CPU usage: 12109 Xorg 58.3%; 3 java 11.6%, 9.3%, 7.6%; soffice.bin 13.2%.

After letting everything sit idle again for about 7 hours, CPU usage: 12109 Xorg 57%; 3 java nowhere in sight (apps terminated by server); soffice.bin 12.6%; FF 5.3%; plugin-container 2.3%.

The last step for today is to kill those spreadsheets and see what happens, CPU usage: 12109 Xorg 1%; soffice.bin not in sight; FF 4.6%; plugin-container 2%.

The disappearance of java in the last two captures is due to disconnect from the application server.

My preliminary conclusion is there's something in java and/or libreoffice that begins causing load on the X server once activity begins. The full effects of the java applet in FF could not be determined during the attempt to duplicate due to the lack of a data feed Saturday and Sunday.

There's apparently something somewhere that holds the X server load high even when the causative items are just sitting idle for extended periods.

The Ksnapshot region selection is still clunky (tested again after this latest closing of the all spreadsheets and libreoffice) with noticeable lag to cursor movement and jerky progression of the window borders, as compared to CentOS pre-6.7 update.

Bill