Recent Posts

Pages: [1] 2 3 ... 10
1
SERT / Re: director_out.txt file
« Last post by JohannPais on May 29, 2020, 08:50:52 PM »
Thank you, Darwin, for sharing your intermittent symptom, which appears to be a new observation -- I recommend (you open) a support ticket with the SERT Support Request Form https://www.spec.org/sert/feedback/issuereport.html, attaching relevant forensic evidence per 7. General Troubleshooting via the following guidance.

2
SERT / director_out.txt file
« Last post by DarwinSmith on May 27, 2020, 09:12:30 AM »
I was wondering if anyone else has seen this behavior.  Once in a while we get a director_out.txt file that has the time-stamped entries out of order. When this happens there are Java exceptions in the file as well.

Some examples of exceptions:
WARNING: Exception during interval
  Caused by: java.io.IOException: Connection reset by peer
WARNING: Exception while ending sequence
WARNING: Exception while ending phase
  Caused by: java.io.IOException: Broken pipe
WARNING: Exception while ending worklet
  Caused by: java.io.IOException: Broken pipe
WARNING: Exception while ending suite
  Caused by: java.io.IOException: Broken pipe
SEVERE: Exception generating reports
SEVERE: Exception while running the suite
3
SERT / Re: config-all.xml documentation
« Last post by DarwinSmith on May 18, 2020, 03:34:10 PM »
Excellent resources, thanks!
4
SERT / Re: config-all.xml documentation
« Last post by JohannPais on May 06, 2020, 05:58:25 PM »
Darwin -- thank you for inquiring into config-all.xml details, as we note the following comment pertaining to compliant configurations.

<!--##############################################################
            For compliant SERT configurations, only the lines below normally
            need to be edited.
         -->


However, section 7.10.2 of http://spec.org/sert2/SERT-designdocument.pdf references interval-series.

Also, slides 48 & 49 of https://www.spec.org/chauffeur-wdk/docs/chauffeurPresentation1.pdf include additional details.
5
PTDaemon / Re: Yokogawa WT3000E
« Last post by GregDarnell on April 23, 2020, 12:28:07 PM »
SPEC last worked on the WT3000 in December 2009 and we were unable to get the device to pass our acceptance tests.  Any results captured using that device with SPEC software would be unreliable, and I would recommend against its use.

You can get a list of device IDs by running the PTDaemon executable with the "-u" (for "unsupported") argument.
6
PTDaemon / Yokogawa WT3000E
« Last post by DarwinSmith on April 23, 2020, 11:56:30 AM »
Is active work being done to support the Yokogawa WT3000E using the Ethernet port?  I saw it in the pick list for the SERTUI with a not that it is experimental.  However, I don't see it in the power measurement setup guide. Does it have a PTD-ID assigned to it?
7
SERT / config-all.xml documentation
« Last post by DarwinSmith on April 23, 2020, 11:05:51 AM »
Does anyone have documentation on the config-all.xml file? For example, what are the options for <interval-series> and what do they mean.
8
Ito ay isang forum. Tindahan ng Kaalaman Hindi ko alam kung saan hahanapin ito.
9
Hi Yvonne,

What this is saying is that 100% of your samples had an uncertainty greater than 1%. It would appear that the current range setting for the idle worket is set too high for the interval. Try editing the range setting for idle in the SERT GUI power analyzer options or by editing listners.xml. It should be set to the average value reported during test, or in this instance what you see on the analyzer when there is no activity on the SUT.

More information can be found at https://www.spec.org/power/docs/SPEC-Power_Measurement_Setup_Guide.pdf

The following is an excerpt from section 2.6:
"For compliant SPECpowerbenchmark runs, specific uncertainty requirements have to be guaranteed according to  the SPECpower  Run  and  Reporting  Rules.  The  uncertainty  of  a measured  power value depends on how well the configured volt and ampere ranges fit to this value. The formulas to calculate the uncertainty often directly include the configured volt and ampere ranges. Selecting the volt range is quite easy as the voltage depends only on the power source and does not change during the test. Selecting the ampere range is more complex, because it differs with the SUT’s power  consumption  during  the  benchmark  run.  A  large  difference  between  different  load  levels  may even require individual ampere range settings for each level. PTDaemon  has  the  ability  to  select  the  closest  range  appropriate  for  a  requested  value.    The information is maintained on a per-analyzer basis.  For example, if you specify a range of 2.2A and the analyzer  in  use  has  ranges  of  0.5,  1.0,  2.0  and  5.0A,  PTDaemon  will  set  the  analyzer  to  the  5.0A range."
10
SERT / NVALID: [YokogawaWT310 (power_localhost_8890_0)] Idle:Idle:idle: The power analy
« Last post by Yvonne on February 25, 2020, 01:29:22 AM »
Hi:
My test result got a problem, it's show"INVALID: [YokogawaWT310 (power_localhost_8890_0)] Idle:Idle:idle: The power analyzer had 100.0% samples with uncertainty of more than 1%, exceeding the threshold of 5.0%" , can you give me some suggestion , thanks so much.
Pages: [1] 2 3 ... 10