|
dev
newsgroups
|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
Duration values misreported in ProfilerProfiler (this is SQL Server 2005 SP2 with rollups) is reporting some
interesting duration values for stored procedure execution: StartTime: 2007-11-01 10:17:58.180. EndTime: 2007-11-01 10:17:58.180. Duration: 18446744073709532 I wonder if anyone else has seen this, or is my server just doing a bit of time travelling? Thank you. DVC Exactly which rollups do you have? I have some minor hot fixes and have
never noticed this behavior. -- Show quoteAndrew J. Kelly SQL MVP Solid Quality Mentors "dvc" <d**@discussions.microsoft.com> wrote in message news:F34C874F-B9FE-468D-89F0-79E4B73C10C0@microsoft.com... > Profiler (this is SQL Server 2005 SP2 with rollups) is reporting some > interesting duration values for stored procedure execution: > StartTime: 2007-11-01 10:17:58.180. > EndTime: 2007-11-01 10:17:58.180. > Duration: 18446744073709532 > > I wonder if anyone else has seen this, or is my server just doing a bit of > time travelling? > > Thank you. > > DVC The server is at 9.00.3054.00, so nothing unusual there, I would have thought.
Show quote "Andrew J. Kelly" wrote: > Exactly which rollups do you have? I have some minor hot fixes and have > never noticed this behavior. > > -- > Andrew J. Kelly SQL MVP > Solid Quality Mentors > > > "dvc" <d**@discussions.microsoft.com> wrote in message > news:F34C874F-B9FE-468D-89F0-79E4B73C10C0@microsoft.com... > > Profiler (this is SQL Server 2005 SP2 with rollups) is reporting some > > interesting duration values for stored procedure execution: > > StartTime: 2007-11-01 10:17:58.180. > > EndTime: 2007-11-01 10:17:58.180. > > Duration: 18446744073709532 > > > > I wonder if anyone else has seen this, or is my server just doing a bit of > > time travelling? > > > > Thank you. > > > > DVC > > That is what I run as well and I have not noticed this behavior. It would be
interesting to see if it can be repeated or even happens again. -- Show quoteAndrew J. Kelly SQL MVP Solid Quality Mentors "dvc" <d**@discussions.microsoft.com> wrote in message news:3F356125-5D4F-4CBC-B7CF-3427206CF5EC@microsoft.com... > The server is at 9.00.3054.00, so nothing unusual there, I would have > thought. > > "Andrew J. Kelly" wrote: > >> Exactly which rollups do you have? I have some minor hot fixes and have >> never noticed this behavior. >> >> -- >> Andrew J. Kelly SQL MVP >> Solid Quality Mentors >> >> >> "dvc" <d**@discussions.microsoft.com> wrote in message >> news:F34C874F-B9FE-468D-89F0-79E4B73C10C0@microsoft.com... >> > Profiler (this is SQL Server 2005 SP2 with rollups) is reporting some >> > interesting duration values for stored procedure execution: >> > StartTime: 2007-11-01 10:17:58.180. >> > EndTime: 2007-11-01 10:17:58.180. >> > Duration: 18446744073709532 >> > >> > I wonder if anyone else has seen this, or is my server just doing a bit >> > of >> > time travelling? >> > >> > Thank you. >> > >> > DVC >> >> Running the trace again, it's continuing to misreport. For example,
StartTime: 2007-11-01 14:52:17.153 EndTime: 2007-11-01 14:52:17.170 Duration: 18446744073709539 It's pretty much arbitrary between stored procedures and times.. This is a production box, so I'm not keen to reboot it to fix what may just be curiosity, but I'll update this discussion as and when I have more information. dvc Show quote "Andrew J. Kelly" wrote: > That is what I run as well and I have not noticed this behavior. It would be > interesting to see if it can be repeated or even happens again. > > -- > Andrew J. Kelly SQL MVP > Solid Quality Mentors > > > "dvc" <d**@discussions.microsoft.com> wrote in message > news:3F356125-5D4F-4CBC-B7CF-3427206CF5EC@microsoft.com... > > The server is at 9.00.3054.00, so nothing unusual there, I would have > > thought. > > > > "Andrew J. Kelly" wrote: > > > >> Exactly which rollups do you have? I have some minor hot fixes and have > >> never noticed this behavior. > >> > >> -- > >> Andrew J. Kelly SQL MVP > >> Solid Quality Mentors > >> > >> > >> "dvc" <d**@discussions.microsoft.com> wrote in message > >> news:F34C874F-B9FE-468D-89F0-79E4B73C10C0@microsoft.com... > >> > Profiler (this is SQL Server 2005 SP2 with rollups) is reporting some > >> > interesting duration values for stored procedure execution: > >> > StartTime: 2007-11-01 10:17:58.180. > >> > EndTime: 2007-11-01 10:17:58.180. > >> > Duration: 18446744073709532 > >> > > >> > I wonder if anyone else has seen this, or is my server just doing a bit > >> > of > >> > time travelling? > >> > > >> > Thank you. > >> > > >> > DVC > >> > >> > > Can you check what are your Power Options on that server (in Windows)? If it
is anything else than 'Always On', it may be causing problems. Time calculations for the duration value are done using the processor frequency as parameter. With power saving on, frequency changes to save energy and makes our calculations inaccurate. Also, when there is more than 1 processor, power savings is applied differently depending on the processor's workload and will cause the values to go out of sync between processors. There were some fixes done in that area for SP2 that eliminated some class of problems, but not completely for situations like above. Regards, Maciek Sarnowicz Show quote "dvc" <d**@discussions.microsoft.com> wrote in message news:CE6CF146-1088-482E-9E77-9963FCCE85D4@microsoft.com... > Running the trace again, it's continuing to misreport. For example, > > StartTime: 2007-11-01 14:52:17.153 > EndTime: 2007-11-01 14:52:17.170 > Duration: 18446744073709539 > > It's pretty much arbitrary between stored procedures and times.. This is a > production box, so I'm not keen to reboot it to fix what may just be > curiosity, but I'll update this discussion as and when I have more > information. > > dvc > > "Andrew J. Kelly" wrote: > >> That is what I run as well and I have not noticed this behavior. It would >> be >> interesting to see if it can be repeated or even happens again. >> >> -- >> Andrew J. Kelly SQL MVP >> Solid Quality Mentors >> >> >> "dvc" <d**@discussions.microsoft.com> wrote in message >> news:3F356125-5D4F-4CBC-B7CF-3427206CF5EC@microsoft.com... >> > The server is at 9.00.3054.00, so nothing unusual there, I would have >> > thought. >> > >> > "Andrew J. Kelly" wrote: >> > >> >> Exactly which rollups do you have? I have some minor hot fixes and >> >> have >> >> never noticed this behavior. >> >> >> >> -- >> >> Andrew J. Kelly SQL MVP >> >> Solid Quality Mentors >> >> >> >> >> >> "dvc" <d**@discussions.microsoft.com> wrote in message >> >> news:F34C874F-B9FE-468D-89F0-79E4B73C10C0@microsoft.com... >> >> > Profiler (this is SQL Server 2005 SP2 with rollups) is reporting >> >> > some >> >> > interesting duration values for stored procedure execution: >> >> > StartTime: 2007-11-01 10:17:58.180. >> >> > EndTime: 2007-11-01 10:17:58.180. >> >> > Duration: 18446744073709532 >> >> > >> >> > I wonder if anyone else has seen this, or is my server just doing a >> >> > bit >> >> > of >> >> > time travelling? >> >> > >> >> > Thank you. >> >> > >> >> > DVC >> >> >> >> >> >> > Can you check what are your Power Options on that server (in Windows)? If it is anything else than Ahh... Does this contributes to below type of messages in the errorlog?> 'Always On', it may be causing problems. "The time stamp counter of CPU on scheduler id 1 is not synchronized with other CPUs." -- Show quoteTibor Karaszi, SQL Server MVP http://www.karaszi.com/sqlserver/default.asp http://sqlblog.com/blogs/tibor_karaszi "Maciek Sarnowicz [MSFT]" <mac***@online.microsoft.com> wrote in message news:342584D6-DB36-4B38-B8CA-DB344D23E08B@microsoft.com... > Can you check what are your Power Options on that server (in Windows)? If it is anything else than > 'Always On', it may be causing problems. > > Time calculations for the duration value are done using the processor frequency as parameter. With > power saving on, frequency changes to save energy and makes our calculations inaccurate. Also, > when there is more than 1 processor, power savings is applied differently depending on the > processor's workload and will cause the values to go out of sync between processors. > > There were some fixes done in that area for SP2 that eliminated some class of problems, but not > completely for situations like above. > > Regards, > Maciek Sarnowicz > > "dvc" <d**@discussions.microsoft.com> wrote in message > news:CE6CF146-1088-482E-9E77-9963FCCE85D4@microsoft.com... >> Running the trace again, it's continuing to misreport. For example, >> >> StartTime: 2007-11-01 14:52:17.153 >> EndTime: 2007-11-01 14:52:17.170 >> Duration: 18446744073709539 >> >> It's pretty much arbitrary between stored procedures and times.. This is a >> production box, so I'm not keen to reboot it to fix what may just be >> curiosity, but I'll update this discussion as and when I have more >> information. >> >> dvc >> >> "Andrew J. Kelly" wrote: >> >>> That is what I run as well and I have not noticed this behavior. It would be >>> interesting to see if it can be repeated or even happens again. >>> >>> -- >>> Andrew J. Kelly SQL MVP >>> Solid Quality Mentors >>> >>> >>> "dvc" <d**@discussions.microsoft.com> wrote in message >>> news:3F356125-5D4F-4CBC-B7CF-3427206CF5EC@microsoft.com... >>> > The server is at 9.00.3054.00, so nothing unusual there, I would have >>> > thought. >>> > >>> > "Andrew J. Kelly" wrote: >>> > >>> >> Exactly which rollups do you have? I have some minor hot fixes and have >>> >> never noticed this behavior. >>> >> >>> >> -- >>> >> Andrew J. Kelly SQL MVP >>> >> Solid Quality Mentors >>> >> >>> >> >>> >> "dvc" <d**@discussions.microsoft.com> wrote in message >>> >> news:F34C874F-B9FE-468D-89F0-79E4B73C10C0@microsoft.com... >>> >> > Profiler (this is SQL Server 2005 SP2 with rollups) is reporting some >>> >> > interesting duration values for stored procedure execution: >>> >> > StartTime: 2007-11-01 10:17:58.180. >>> >> > EndTime: 2007-11-01 10:17:58.180. >>> >> > Duration: 18446744073709532 >>> >> > >>> >> > I wonder if anyone else has seen this, or is my server just doing a bit >>> >> > of >>> >> > time travelling? >>> >> > >>> >> > Thank you. >>> >> > >>> >> > DVC >>> >> >>> >> >>> >>>  |
|||||||||||||||||||||||
Other interesting topics