You should really only be running heap at 50% of total system RAM, that gives the other half to allow your system to cache FS data and improve performance.
However you may find that moving to more smaller nodes will push your use a little higher, so you may end up needing more nodes as a result. Regards, Mark Walkom Infrastructure Engineer Campaign Monitor email: [email protected] web: www.campaignmonitor.com On 26 May 2014 14:59, Ran Franco <[email protected]> wrote: > Ho sorry.. > Its get around 90%-95% > > On Monday, May 26, 2014 7:50:43 AM UTC+3, Mark Walkom wrote: >> >> But what is the use, are you close to 100% use of that? >> >> Regards, >> Mark Walkom >> >> Infrastructure Engineer >> Campaign Monitor >> email: [email protected] >> web: www.campaignmonitor.com >> >> >> On 26 May 2014 14:45, Ran Franco <[email protected]> wrote: >> >>> Hi Mark, >>> >>> My current heap size 12GB >>> >>> On Monday, May 26, 2014 2:01:41 AM UTC+3, Mark Walkom wrote: >>>> >>>> It depends. You probably won't get a linear drop in heap size by >>>> halving RAM and amount of data on a server. >>>> What's your current heap use like now? >>>> >>>> Regards, >>>> Mark Walkom >>>> >>>> Infrastructure Engineer >>>> Campaign Monitor >>>> email: [email protected] >>>> web: www.campaignmonitor.com >>>> >>>> >>>> On 25 May 2014 21:36, Ran Franco <[email protected]> wrote: >>>> >>>>> Hi all, >>>>> >>>>> Currently I'm using ES cluster with 8 nodes (m3.xlarge machines from >>>>> Amazon) >>>>> These nodes contain - >>>>> Ram - 15.0 GB >>>>> CPU - 13 (4 core x 3.25 unit) >>>>> >>>>> I wanted to know if i'll get better performance whether I'll use 16 >>>>> nodes but weaker (m3.large) >>>>> Ram - 7.5 GB >>>>> CPU - 6.5 (2 core x 3.25 unit) >>>>> >>>>> What is preferred? a bigger cluster with weaker machines? or smaller >>>>> cluster with stronger machines? >>>>> >>>>> >>>>> Thanks ! >>>>> Ran >>>>> >>>>> -- >>>>> You received this message because you are subscribed to the Google >>>>> Groups "elasticsearch" group. >>>>> To unsubscribe from this group and stop receiving emails from it, send >>>>> an email to [email protected]. >>>>> To view this discussion on the web visit https://groups.google.com/d/ >>>>> msgid/elasticsearch/1ebe28b7-fa45-495f-94d8-400976b75ae5%40goo >>>>> glegroups.com<https://groups.google.com/d/msgid/elasticsearch/1ebe28b7-fa45-495f-94d8-400976b75ae5%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>> . >>>>> For more options, visit https://groups.google.com/d/optout. >>>>> >>>> >>>> -- >>> You received this message because you are subscribed to the Google >>> Groups "elasticsearch" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> To view this discussion on the web visit https://groups.google.com/d/ >>> msgid/elasticsearch/9b73d836-8bb8-44c0-81c9-709038f70629% >>> 40googlegroups.com<https://groups.google.com/d/msgid/elasticsearch/9b73d836-8bb8-44c0-81c9-709038f70629%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- > You received this message because you are subscribed to the Google Groups > "elasticsearch" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/elasticsearch/1cef7433-2414-4c45-bcfe-eaf65e9cbc2e%40googlegroups.com<https://groups.google.com/d/msgid/elasticsearch/1cef7433-2414-4c45-bcfe-eaf65e9cbc2e%40googlegroups.com?utm_medium=email&utm_source=footer> > . > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "elasticsearch" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CAEM624atd2e%2BRD%2BJpm31YNtvsKcdtNsDWSmtKjoTCYixoRE9%2Bg%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
