September 16, 2016 at 2:51 am
we are getting alerts from SCOM that
page life expectancy of sql instance" "mssqlserver on computer is too low, how can we solve and trouble shoot the issue ?
September 16, 2016 at 2:56 am
What are the thresholds?
September 16, 2016 at 3:14 am
What does the PLE look like over a day or more?
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
September 16, 2016 at 5:41 am
Are you experiencing any other issues related to memory or disk? What are your top wait statistics? Are you experiencing performance issues. PLE, all by itself, doesn't really indicate much of anything within SQL Server.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
September 16, 2016 at 7:22 am
I honestly haven't used PLE as a metric in probably a decade. Avg disk sec/read and avg disk sec/write are WAY more appropriate IMHO.
Best,
Kevin G. Boles
SQL Server Consultant
SQL MVP 2007-2012
TheSQLGuru on googles mail service
September 16, 2016 at 7:29 am
TheSQLGuru (9/16/2016)
I honestly haven't used PLE as a metric in probably a decade. Avg disk sec/read and avg disk sec/write are WAY more appropriate IMHO.
While obviously not expecting you to give away your living, if you have a link to anything you may have written on this I'd be very interested
I'm a DBA.
I'm not paid to solve problems. I'm paid to prevent them.
September 16, 2016 at 10:13 am
andrew gothard (9/16/2016)
TheSQLGuru (9/16/2016)
I honestly haven't used PLE as a metric in probably a decade. Avg disk sec/read and avg disk sec/write are WAY more appropriate IMHO.While obviously not expecting you to give away your living, if you have a link to anything you may have written on this I'd be very interested
I don't blog, don't have a website, don't even keep a resume up to date. I somehow still seem to get as much work as I want. 😎
Think of it this way: You have a 64GB server and PLE is 300 (the number often referenced by various online resources IIRC). That means your server is turning over ~60GB of data every 5 minutes. That could be just fine or REALLY bad, depending largely on the capability of your IO subsystem. So why not just measure and be concerned with the performance of the IO subsystem? And that is primarily the 2 metrics I mentioned (tracked automatically by the engine per database file, btw). Either you are waiting on physical IO or you aren't.
Best,
Kevin G. Boles
SQL Server Consultant
SQL MVP 2007-2012
TheSQLGuru on googles mail service
September 16, 2016 at 10:28 am
Just to throw my $0.02 in, I definitely agree that seeing low PLE does not necessarily mean there's a problem. As Kevin and others have mentioned, if the performance of the storage subsystem faced with all that physical IO is acceptable to the business, then there may not be a problem worth fixing.
Having said that, though, a well-performing array, even if all flash, is going to be significantly slower than memory. It really comes down to a question of whether the performance improvement from increasing PLE is worth it to the business.
Even that's not the full answer, though, since while a read from memory will be faster than a read from a well-performing array, if physical IO is your bottleneck and you bump up memory to accommodate the IO, you may move the bottleneck somewhere more painful.
A better approach (than just throwing memory at it) would be to see what is driving all that IO and seeing if you could reduce the IO required for those processes. You still might end up moving the bottleneck somewhere more painful, though, although it's less likely in this case.
"It depends..." in spades 🙂
Cheers!
Viewing 8 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply