Hi, 

I have few servers that are VM and different version of SQL Severs (2008 Express/Standard, 2012, or 2014 Enterprise) and seems all of them have same issue that when I am querying "sys.dm_os_performance_counters" for "[object_name] LIKE '%Buffer Manager%' AND [counter_name] = 'Page life expectancy'" I am getting very big number (Ex, above 100K or some servers 1M). 

Is this number seems fine or acceptable?

I believe I found my answer but need someone else to confirm If I am correct.

The number is cumulative and If you want to get real number you have to do sampling, meaning that you have to create a job and then capture that information every 10, 15 minutes and compare numbers either with MS suggested 300 or Jonathan Kehayias method (DataCacheSizeInGB/4GB*300) in "https://www.sqlskills.com/blogs/jonathan/finding-what-queries-in-the-plan-cache-use-a-specific-index/"

'Page life expectancy'" I am getting very big number

Hello Arbi,

The rule for this performance counter value: The higher the better.

The counter is the time span how long in average the buffer pages (=cache) stays in memory, before it's swapped out to release space for other data pages. So a value of 100k and more is a good value and indicates that SQL Server can often and for long use data pages for buffer pool instead of reading them from disk (which is slow, of co

Is this number seems fine or acceptable?

@Olaf & @Shanky:

If I capture every 15 minutes PLE and then compare the results against 300 or any number like (DataCacheSizeInGB/4GB*300) is that going to give me more realistic number

2015-05-27 09:45:01.603 --> 144598

2015-05-27 10:00:01.093 --> 145497

Difference --> 899 

THEN Compare 899 vs 300 

@Olaf & @Shanky:

If I capture every 15 minutes PLE and then compare the results against 300 or any number like (DataCacheSizeInGB/4GB*300) is that going to give me more realistic number

2015-05-27 09:45:01.603 --> 144598

2015-05-27 10:00:01.093 --> 145497

Difference --> 899 

THEN Compare 899 vs 300