in reply to CPAN.pm always killed on VPS

You may be running out of memory. I'm pretty sure CPAN.pm slurps that whole file into a single data structure.

Replies are listed 'Best First'.
Re^2: CPAN.pm always killed on VPS
by halfcountplus (Hermit) on Jul 21, 2009 at 03:56 UTC
    You're probably right. I've never dealt with this before! But right after the kill /var/log/messages gets littered with a bunch of (incomprehensible to me) memory stuff, starting with
    Jul 21 03:46:47 localhost kernel: oom-killer: gfp_mask=0x280d2, order= +0 Jul 21 03:46:47 localhost kernel: Jul 21 03:46:47 localhost kernel: Call Trace: Jul 21 03:46:47 localhost kernel: [<ffffffff80258cd0>] out_of_memory+0 +x2e/0x182
    I guess that's not so incomprehensible =/ Thanks!

    Probably there is no way out of that...

      I haven't used CPAN::SQLite, but its intent is to reduce the amount of memory required.

      Update: I've just enabled it per the documentation; the CPAN shell uses a third the memory that it did before.

        Hmmm. I was all excited for a minute there and then:
        Fetching with LWP: http://www.cpan.org/modules/03modlist.data.gz Creating database file ... Gathering information from index files ... Catching error: "system /usr/bin/perl -MCPAN::SQLite::META qw(setup up +date) -e setup failed: 9 at /usr/local/share/perl/5.10.0/CPAN/SQLite/ +META.pm line 323, <FIN> line 2.\cJ" at /usr/share/perl/5.10/CPAN.pm l +ine 391 CPAN::shell() called at -e line 1 Creating database file ... Gathering information from index files ... system /usr/bin/perl -MCPAN::SQLite::META qw(setup update) -e setup fa +iled: 9 at /usr/local/share/perl/5.10.0/CPAN/SQLite/META.pm line 323, + <FIN> line 2.
        Some extraneous stuff culled.