[hpsdr] Excessive Page Faults

[Dick Bingham]

Hello Bill ( KD5TFD)

You asked ===>
"* Message: 2*





















*Date: Sun, 24 May 2020 22:36:06 -0500From: Bill Tracey <bill at ewjt.com
<bill at ewjt.com>>To: Dick Bingham <dick.bingham at gmail.com
<dick.bingham at gmail.com>>,hpsdr at lists.openhpsdr.org
<hpsdr at lists.openhpsdr.org>Subject: Re: [hpsdr] Excessive Number of Page
Faults While Running        PowerHPSDR mRX PS V3.4.7Message-ID:
<1590377965604.0e64f70564411266 at ewjt.com
<1590377965604.0e64f70564411266 at ewjt.com>>Content-Type: text/plain;
charset="iso-8859-1"; format=flowedNot sure why you're homing in on page
faults - Is47E6 page faults in 8 hours excessive?  Where areyou getting
this data - task manager?  What doesPF delta look like?   Whose page faults
are these- OpenHPSDR mRx of WSJT-X?   How much memory doyou have?  If it's
page faulting like crazy youmay be thrashing due to insufficient memory
tohold the working set?  Is this a new behavior -had this been working?
 If so, what's changedbetween the working time and non working
time?Regards,Bill (kd5tfd)* "

Well, *homing in* on page faults seems like a good place to wonder about
what is
happening given the - now - 289-million for the top faulter way exceeding
the next
highest contender at a bit over 1-million.

As of this writing the PF counter is up to 289.xxxE6 faults after running
for about 8-hours.
The *PF-Delta *is running at a nearly steady "4" but I see excursions
periodically in the
7000 range. These numbers are as presented on *Task Manager.*

These numbers are for the OpenHPSDR mRX PS program ===>
[image: image.png]
This is pretty small print so you may not be able to read it. The top line
for
*PowerSDR.exe *reads:
=================
CPU ===> 10/15%
Working Set (Memory) ===> *199,550K*
Peak Working Set (Memory) ===> *237,548K*
Working Set Delta ===> zero K
Memory Private ===> *164,380K*
Commit Size ===> 215,468K
Paged Pool ===> 403K
NP Pool ===> 41K
Page Faults ===> 289,474,960K
PF Delta ===> 4 ===>7 ===> BIG number up to 7K at times

The next line is for *JT9* and its *PF* is relatively small at 1,105,000K
with *PF-Delta* at zero
================================================================

I have close to 4GB of memory and as time progresses "Leakage" or whatever
is consuming
available memory to the point my computer will not allow *Task Manager to
start *due to
'insufficient resources'. At this point I have to shut down the computer
and restart it to have
a functioning machine.

This is 'new' behavior that began several weeks ago (I had changed nothing
in my computer
for months) when *OpenHPSDR mRX PS* screen display stopped working during
the receive
cycle of a CQ session on FT8. I would then cycle the power button to
reactivate the RX only
to have the lock-up happen again at what seems like a random basis

How do I increase the memory 'given' to running the OpenHPSDR to ensure the
'thrashing'
you described is not an issue?

Thank you for your response to my question - this old RF and Microwave
Frequencies Engineer
is out of his league in things-Computer Science.

73  Dick/w7wkr
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20200525/51f5ea22/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 86411 bytes
Desc: not available
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20200525/51f5ea22/attachment-0001.png>