-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thu 12 Jun 2014 10:18:03 PM UTC, Nicholas Lee via smartos-discuss wrote: > > > > On Fri, Jun 13, 2014 at 1:13 AM, Filip Hajny <[email protected] > <mailto:[email protected]>> wrote: > > Nicholas, > > can you make sure you’re not nearing your memory cap in the VM? > I’ve seen CPAN fail with ENOMEM even when over 200M memory was > available. It’s a huge beast. > > -F > > > Looks like that may have been the problem. Sounds like it is a huge > beast as the vm had 512Mb. Maybe not very well optimised for 64bit. > > I've been busy with other stuff, so haven't run any thing on this node > since the last attempt. Only got 8Gb on this system - it's pretty old > - will have to look at getting my new hardware sorted. > > [root@00-15-17-20-9d-de ~]# vmadm get > 94391c9b-32df-4295-9e8d-24891ed6a03b max_physical_memory | grep memory > "max_physical_memory": 512, > "max_locked_memory": 512, > > > [root@00-15-17-20-9d-de ~]# vmadm update > 94391c9b-32df-4295-9e8d-24891ed6a03b max_physical_memory=2048 > Successfully updated VM 94391c9b-32df-4295-9e8d-24891ed6a03b > [root@00-15-17-20-9d-de ~]# vmadm get > 94391c9b-32df-4295-9e8d-24891ed6a03b max_physical_memory | grep memory > "max_physical_memory": 2048, > "max_locked_memory": 2048, > [root@00-15-17-20-9d-de ~]# vmadm list | grep kwww > 94391c9b-32df-4295-9e8d-24891ed6a03b OS 2048 running > kwww2 > [root@00-15-17-20-9d-de ~]# zlogin 94391c9b-32df-4295-9e8d-24891ed6a03b > [Connected to zone '94391c9b-32df-4295-9e8d-24891ed6a03b' pts/2] > Last login: Thu Jun 12 11:46:32 on pts/2 > __ . . > _| |_ | .-. . . .-. :--. |- > |_ _| ;| || |(.-' | | | > |__| `--' `-' `;-| `-' ' ' `-' > / ; Instance (base64 14.1.0) > `-' http://wiki.joyent.com/jpc2/SmartMachine+Base > > [root@kwww2 ~]# sm-meminfo rss > * Gathering rss memory infomation for > 94391c9b-32df-4295-9e8d-24891ed6a03b.. > Memory (RSS) Cap 2048M > Memory (RSS) Used 111M > Memory (RSS) Free 1936M > Memory NOver Cap 0 > Memory Total Pgout 0M > [root@kwww2 ~]# cpan > Terminal does not support AddHistory. > > cpan shell -- CPAN exploration and modules installation (v2.05) > Enter 'h' for help. > > cpan[1]> install Digest::MD5 > Reading '/root/.cpan/Metadata' > Database was generated on Wed, 11 Jun 2014 16:41:02 GMT > Fetching with LWP: > http://mirror.webtastix.net/CPAN/authors/01mailrc.txt.gz > Reading '/root/.cpan/sources/authors/01mailrc.txt.gz' > ............................................................................DONE > Fetching with LWP: > http://mirror.webtastix.net/CPAN/modules/02packages.details.txt.gz > Reading '/root/.cpan/sources/modules/02packages.details.txt.gz' > Database was generated on Thu, 12 Jun 2014 16:29:02 GMT > ............................................................................DONE > Fetching with LWP: > http://mirror.webtastix.net/CPAN/modules/03modlist.data.gz > Reading '/root/.cpan/sources/modules/03modlist.data.gz' > DONE > Writing /root/.cpan/Metadata > Running install for module 'Digest::MD5' > Checksum for > /root/.cpan/sources/authors/id/G/GA/GAAS/Digest-MD5-2.53.tar.gz ok > Scanning cache /root/.cpan/build for sizes > ............................................................................DONE > Configuring G/GA/GAAS/Digest-MD5-2.53.tar.gz with Makefile.PL > Perl's config says that U32 access must be aligned. > Checking if your kit is complete... > Looks good > Writing Makefile for Digest::MD5 > Writing MYMETA.yml and MYMETA.json > GAAS/Digest-MD5-2.53.tar.gz > /opt/local/bin/perl Makefile.PL -- OK > Running make for G/GA/GAAS/Digest-MD5-2.53.tar.gz > cp MD5.pm blib/lib/Digest/MD5.pm > /opt/local/bin/perl /opt/local/lib/perl5/5.18.0/ExtUtils/xsubpp > -typemap /opt/local/lib/perl5/5.18.0/ExtUtils/typemap -typemap > typemap MD5.xs > MD5.xsc && mv MD5.xsc MD5.c > gcc -c -D_REENTRANT -O2 -pipe -O2 -pthread -I/opt/local/include > -I/usr/include -fno-strict-aliasing -I/opt/local/include -m64 > -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -DPERL_USE_SAFE_PUTENV -O2 > -pipe -O2 -pthread -I/opt/local/include -I/usr/include > -DVERSION=\"2.53\" -DXS_VERSION=\"2.53\" -fPIC > "-I/opt/local/lib/perl5/5.18.0/x86_64-solaris-thread-multi-64/CORE" > -DU32_ALIGNMENT_REQUIRED MD5.c > Running Mkbootstrap for Digest::MD5 () > chmod 644 MD5.bs > rm -f blib/arch/auto/Digest/MD5/MD5.so > gcc -shared -m64 > -L/opt/local/gcc47/lib/gcc/x86_64-sun-solaris2.11/4.7.3 > -Wl,-R/opt/local/gcc47/lib/gcc/x86_64-sun-solaris2.11/4.7.3 > -L/opt/local/gcc47/lib -Wl,-R/opt/local/gcc47/lib -L/opt/local/lib > -L/usr/lib/amd64 -Wl,-R/usr/lib/amd64 -Wl,-R/opt/local/lib MD5.o -o > blib/arch/auto/Digest/MD5/MD5.so > chmod 755 blib/arch/auto/Digest/MD5/MD5.so > cp MD5.bs blib/arch/auto/Digest/MD5/MD5.bs > chmod 644 blib/arch/auto/Digest/MD5/MD5.bs > Manifying blib/man3/Digest::MD5.3 > GAAS/Digest-MD5-2.53.tar.gz > /opt/local/bin/bmake -- OK > Running make test > PERL_DL_NONLAZY=1 /opt/local/bin/perl "-MExtUtils::Command::MM" "-e" > "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t > t/align.t .... ok > t/badfile.t .. ok > t/bits.t ..... ok > t/clone.t .... ok > t/files.t .... ok > t/md5-aaa.t .. ok > t/threads.t .. ok > t/utf8.t ..... ok > All tests successful. > Files=8, Tests=277, 1 wallclock secs ( 0.09 usr 0.05 sys + 0.28 > cusr 0.16 csys = 0.59 CPU) > Result: PASS > GAAS/Digest-MD5-2.53.tar.gz > /opt/local/bin/bmake test -- OK > Running make install > Files found in blib/arch: installing files in blib/lib into > architecture dependent library tree > Installing > /opt/local/lib/perl5/site_perl/5.18.0/x86_64-solaris-thread-multi-64/auto/Digest/MD5/MD5.bs > Installing > /opt/local/lib/perl5/site_perl/5.18.0/x86_64-solaris-thread-multi-64/auto/Digest/MD5/MD5.so > Installing > /opt/local/lib/perl5/site_perl/5.18.0/x86_64-solaris-thread-multi-64/Digest/MD5.pm > Installing /opt/local/lib/perl5/site_perl/man/man3/Digest::MD5.3 > Appending installation info to > /opt/local/lib/perl5/5.18.0/x86_64-solaris-thread-multi-64/perllocal.pod > GAAS/Digest-MD5-2.53.tar.gz > /opt/local/bin/bmake install -- OK > > cpan[2]> >
Hi Nicholas: One thing to watch out for is SmartOS defaults to 1:1 swap to max_physical_memory. It seems like it's better to have 2:1 since /tmp uses swap as well. Every JPC (Joyent Public Cloud) package has a 2:1 for swap to memory. I've started explicitly specifying swap in my vmadm create JSON payloads. I got bitten by this with the GHC Builder. Best, Alain -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBAgAGBQJTmjfGAAoJEP0rIXJNjNSAVLwH/izL/qYtPnb5WJbHOGdSjivA cuZ2dTgXqnZ0ggJ9Q6I4kzKslYXz021tgbWUJkTmqrJTw0esnDKdKfOd9Y8zvNVz EachSetDCvlbxHBDbCToMQO2jneGdNon7etRVOJvSeVjIwCQgekB4//Bjw1h3+9r +b2QWLgz5kPYoHCUAiOqvyzb/eOAb3xS0dOiuZLVs2KVh+IJ0sFqQbUrTt9t4ufu r2jsSKkCZ3pzdDEXbFm9qa+912WNI/0/PMPSAbljdcJ9hX6JTzWjTw11cD7yevo4 V5QDVqRBJwajOj/g5QQrlWbfBYSz07fCwXi1/ZWDKy4Xslt82t7po6HTqCTySFE= =9ozS -----END PGP SIGNATURE----- ------------------------------------------- smartos-discuss Archives: https://www.listbox.com/member/archive/184463/=now RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00 Modify Your Subscription: https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb Powered by Listbox: http://www.listbox.com
