Hi there colleagues, I have stable/10 on a rather big machine (2*8*2 e5 Xeon, 64G RAM, SAS+SSD ZFS raid10+ZIL+L2ARC) acting as a PostgreSQL server. To use such a big resource pool that is mostly idle, I'd deployed poudriere there (using tmpfs) too. Most times this combination works like a charm: LA could be 60+ and no visual latency increase on SQL queries. However, sometimes postgres processes got killed by 'out of swap space'. I suppose the source of problem could be that VSZ size of postgres processes (8-9 G) is bigger than swap congigured (4G). Is there any way to prevent this, besides reallocating space for swap? Quick googling does not help, at least I could not find answers relevant enough. Thanks! -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck at FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck at rinet.ru *** ------------------------------------------------------------------------
Baptiste Daroussin
2015-Apr-25 10:43 UTC
protecting some processes from out-of-swap killer
On Sat, Apr 25, 2015 at 01:31:14PM +0300, Dmitry Morozovsky wrote:> Hi there colleagues, > > I have stable/10 on a rather big machine (2*8*2 e5 Xeon, 64G RAM, SAS+SSD ZFS > raid10+ZIL+L2ARC) acting as a PostgreSQL server. > > To use such a big resource pool that is mostly idle, I'd deployed poudriere > there (using tmpfs) too. > > Most times this combination works like a charm: LA could be 60+ and no visual > latency increase on SQL queries. > > However, sometimes postgres processes got killed by 'out of swap space'. > I suppose the source of problem could be that VSZ size of postgres processes > (8-9 G) is bigger than swap congigured (4G). > > Is there any way to prevent this, besides reallocating space for swap? > > Quick googling does not help, at least I could not find answers relevant > enough. > > Thanks!protect(1) ? Best regards, Bapt -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: not available URL: <http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20150425/543070e3/attachment.sig>
Konstantin Belousov
2015-Apr-25 10:47 UTC
protecting some processes from out-of-swap killer
On Sat, Apr 25, 2015 at 01:31:14PM +0300, Dmitry Morozovsky wrote:> Hi there colleagues, > > I have stable/10 on a rather big machine (2*8*2 e5 Xeon, 64G RAM, SAS+SSD ZFS > raid10+ZIL+L2ARC) acting as a PostgreSQL server.64G RAM is not big by modern standards, normally populated desktop can take 32G.> > To use such a big resource pool that is mostly idle, I'd deployed poudriere > there (using tmpfs) too. > > Most times this combination works like a charm: LA could be 60+ and no visual > latency increase on SQL queries. > > However, sometimes postgres processes got killed by 'out of swap space'. > I suppose the source of problem could be that VSZ size of postgres processes > (8-9 G) is bigger than swap congigured (4G). > > Is there any way to prevent this, besides reallocating space for swap?Read protect(1).> > Quick googling does not help, at least I could not find answers relevant > enough. > > Thanks! > > > -- > Sincerely, > D.Marck [DM5020, MCK-RIPE, DM3-RIPN] > [ FreeBSD committer: marck at FreeBSD.org ] > ------------------------------------------------------------------------ > *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck at rinet.ru *** > ------------------------------------------------------------------------ > _______________________________________________ > freebsd-stable at freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-stable > To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"