http://www.windowsitpro.com/authors/author/author/5086760/rsscomment/5086760en-USSun, 27 May 2012 06:04:46 GMTSun, 27 May 2012 06:04:46 GMThttp://www.windowsitpro.com/article/performance/paging-peformance#commentsAnchorTue, 11 Jan 2005 18:50:36 GMT
MY SYSTEM WONT ALLOW ME TO GET INTO WINDOWS BECAUSE MY PAGE FILE IS TO SMALL, CAN YOU HELP ME]]>Anonymous User Tue, 11 Jan 2005 18:50:36 GMThttp://www.windowsitpro.com/article/performance/paging-peformance#commentsAnchorhttp://www.windowsitpro.com/article/task-scheduling/win2k-task-scheduler#commentsAnchorMon, 09 Apr 2001 21:48:07 GMT
This person did not do their homework. "Scheduled tasks" is buggy, unreliable and lost some of the functionality that was there with the old "at" scheduler in NT. For Example, you cannot control it from the command line and if you do any remote management as I’m sure most of us do, it is unusable. Next, if you open/review/modify an existing task, it looses the security context and you must re-enter the username/password, Lastly, Scheduled tasks randomly lose their username/password context information and simply stop running. This is a reported bug as listed on m$ website but you have to pay for a support call to get "the fix" Huh? How much did I pay for this software?!? Should I trust my data to this? ... I think not.]]>Count Ludwig Mon, 09 Apr 2001 21:48:07 GMThttp://www.windowsitpro.com/article/task-scheduling/win2k-task-scheduler#commentsAnchorhttp://www.windowsitpro.com/article/task-scheduling/win2k-task-scheduler#commentsAnchorWed, 07 Feb 2001 01:48:03 GMT
The article does not make it clear that it is impossible to schedule tasks on remote machines without quite some effort, unlike using the "at" command, which can under NT4. Sure you can copy .job files, but they lose their security information when they arrive at the remot machine (though I do see why this should be). This means you must use the GUI tool to set a username and password for the remote job to run. As someone who has been used to being able to schedule remote "at" jobs on hundreds of machines, this is less than useful. An equivalent command line tool to "at" to do remote job submission in Windows 2000 would have been a good idea.]]>Paul Belli Wed, 07 Feb 2001 01:48:03 GMThttp://www.windowsitpro.com/article/task-scheduling/win2k-task-scheduler#commentsAnchorhttp://www.windowsitpro.com/article/performance/paging-peformance#commentsAnchorWed, 11 Aug 1999 10:05:08 GMT
In “Paging Performance” (May), Kenton Gardinier states that a system with multiple hard disks can make simultaneous I/O requests to the various hard disks. This statement is true for SCSI disks, which have queuing, but is it true for EIDE disks, which process only one request at a time? (Hard disks are rarely connected to the secondary EIDE controller, so the primary controller processes all requests for hard disk access.) The author also states that the worst case occurs when the initial and maximum paging file sizes are equal. I purposely set the values equal (and large) to prevent paging file fragmentation. Perhaps the article could have included instructions to address paging file fragmentation (e.g., use Control Panel to put the file temporarily on another partition, run a defragmentation utility, and put the file back).
--Michael Vuister

I agree with your statement regarding the disk subsystem configuration. The article should have included clarification. The paging file fragments under only two conditions: You have a low amount of disk space during the creation of the paging file at startup, or the paging file is forced to grow beyond the minimum value during normal operation. If the paging file needs more space, it dynamically increases as needed (and with a warning) up to the maximum value. After the paging file grows to the maximum value, the system becomes unstable and can crash. In your configuration, you might be wasting disk space and not providing any padding to prevent instability. In other words, what happens to your system if the paging file reaches the minimum or maximum value?
--Kenton Gardinier]]>Michael Vuister Wed, 11 Aug 1999 10:05:08 GMThttp://www.windowsitpro.com/article/performance/paging-peformance#commentsAnchorhttp://www.windowsitpro.com/article/performance/paging-peformance#commentsAnchorWed, 11 Aug 1999 08:38:08 GMT
I read and thoroughly enjoyed Kenton Gardinier’s “Paging Performance” (May) about monitoring and tuning the paging file to improve performance. I’d like to point out that following the advice in the article and dividing the swap file across different disks will prevent the creation of a memory.dmp file in the event of a blue screen trap. Windows NT displays a warning about this situation, which might be acceptable to some users.
--Al Mulnick

NT doesn’t create a memory dump after a system crash in several situations: for example, when the %SYSTEMROOT% partition or other specified path does not have enough space, you haven’t selected the overwrite option, or you haven’t selected an adequate paging file size on the %SYSTEMROOT% partition. In the article, I didn’t mention the correlation between paging file size and memory dumps. To allow memory dumps, you must keep a paging file on the system partition with a size equal to or greater than the amount of physical memory plus an additional MB. Thanks for mentioning this point. --Kenton Gardinier]]>Al Mulnick Wed, 11 Aug 1999 08:38:08 GMThttp://www.windowsitpro.com/article/performance/paging-peformance#commentsAnchor