I just tried it and to me it looks like that 7200000 is some kind of local adjustment for timezone and/or daylight savings time. (7200000 is 2 hrs worth of milliseconds) I tried it with 10984684250000 (which should be 10/22/04 19:07 - corrected for EST and DST) I plugged this into the excel formula and it gave me 10/22/04 20:07. So, I'm not sure what the formula should be, maybe someone else has deciphered it further...
J ----- Original Message ----- From: "Chuck Henson" <[EMAIL PROTECTED]> To: "JMeter Users List" <[EMAIL PROTECTED]> Sent: Friday, October 22, 2004 1:16 PM Subject: RE: How to convert timestamp > Hi Bob.. > What a great piece of information! I've certainly been doing this the hard > way. > However, I cannot seem to get this formula to work. I understand where all > of the values are coming from (except the 7200000 that is added to the > jmeter timestamp). the 86400000 is the number of miliseconds in a day (since > jmeter records it's timestamps in miliseconds), that converts the jmeter > style of total number of miliseconds to Excel style of fractions of a day > and the 25569 is the number of days from 1/1/1900 to 1/1/1970. dividing by > the number of miliseconds in a day and then adding the number of days > between the Excel startpoint and the Unix epoch should do the conversion. > I'm confused about the 7200000 being added to the orginal timestamp, but the > formula doesn't convert to the correct timestamp regardless of whether I add > that value or not.. without the value, it's closer but still off by a couple > of hours (and minutes and seconds). I'm comparing the converted timetamp for > the first sample to the timestamp in the jmeter.log for when the first > thread started. It is possible I suppose that they could be a second (maybe > even two) off, since I assume the jmeter.log timestamp is at the beginning > of the first threads run and the first timestamp in the jtl is at the end of > the first sample, but not several hours. I tried rewriting the formula > simply based on what I learned reading about Excel time values (thanks Joel) > and knowing that the jmeter timestamp is the number of miliseconds since the > epoch which gave me this [(A1/1000)+2209161600)/86400]. here I'm just > converting the timestamp to number of seconds first then adding the number > of seconds from 1/1/1900 to 1/1/1970 then dividing that total by the number > of seconds in a day. this didn't work either and yielded the same result as > the formula you supplied when I removed the mysterious 7200000 from the > equation. > I'm baffled, any suggestions? > Thanks, Chuck > > -----Original Message----- > From: Coret Bob [mailto:[EMAIL PROTECTED] > Sent: Friday, October 22, 2004 9:20 AM > To: JMeter Users List; [EMAIL PROTECTED] > Subject: RE: How to convert timestamp > > > > > I have found at least one way of transforming the JMeter results timestamp > into a date and time using Excel. > > If cell A1 holds the timestamp (eg. 1098446512326) and you put the formula > "=(A1+7200000)/86400000+25569" into cell B1 and change the format in > "d-m-jjjj u:mm" it reads "22-10-2004 14:01". I have scraped the formula > together from several sources, I can't yet explain all the numbers in the > formula... > > Regards, > Bob Coret > > > -----Original Message----- > From: Coret Bob > Sent: Fri 22-10-2004 15:35 > To: [EMAIL PROTECTED] > Cc: > Subject: How to convert timestamp > The sample results contain a field called timeStamp. How (which algorithm) > can I use to convert this timestamp to a format like HH:MM:SS DD-MM-YYYY ? > > I'd like to know the algoritm so I can make the calculation using Perl, XLST > of Excel... > > <sampleResult threadName="Test Scenario 11-1" responseMessage="OK" > timeStamp="1098442834129" dataType="text" label="00 - Startpagina" > responseCode="200" time="93" success="true"/> > > Regards, > Bob Coret > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]