06-09-2016 12:37 AM
We've SAS GRID Manager version 9.4 M3 running on Linux Red Hat 7.
We had memory leak events on the primary node and this problem is fixed after the reboot.
When we see via TOP command on Linux console, all resource are low use.
Can anyone help ?
06-09-2016 04:22 AM
Well that looks like this SAS Note:
Only problem is it says it was fixed in 9.4 M2. I recommend you track this to SAS Tech Support if you haven't done so already.
06-14-2016 11:06 AM
Thank you for reply.
We already have the M3 version.
Our system operation is LINUX RedHat 7.1 wirh VERITAS 6.2
We're looking for a problem on sytem operation tier.
06-16-2016 10:42 AM
When you look at the amount of memory used with the top command, you see the percentage of memory pages "touched". Since SAS does I/O via the operating system's file cache, memory pages become "dirty". top reports all actively used and "dirty" pages in their report.
With a reboot, all the "dirty" pages in memory being used by filec cache will be flushed and therefore not reported as "touched" with top.
Please note that "dirty" pages are free to be used by any application that needs active pages.
Also, 100% usage of memory without swapping is fine. You should only be concerned when you start to see lots of swapping/paging to disk along with 100% usage of memory.
Please let me know if you need further clarification on this.