http://qs321.pair.com?node_id=152724

scottstef has asked for the wisdom of the Perl Monks concerning the following question:

I am trying to install Net::SFTP (via cpan) and one of the requirements is Math::Pari. When I install Math::Pari, it seems to take forever on its testing. Running top on the machine shows the cpu being pegged and it has been like this for over a day.

7:59am up 21:47, 4 users, load average: 1.18, 1.55, 1.71 80 processes: 70 sleeping, 3 running, 2 zombie, 5 stopped CPU states: 98.8% user, 1.1% system, 0.0% nice, 0.0% idle Mem: 126964K av, 123812K used, 3152K free, 41792K shrd, 1562 +4K buff Swap: 205088K av, 41924K used, 163164K free 2222 +8K cached PID USER PRI NI SIZE RSS SHARE STAT LIB %CPU %MEM TIME COM +MAND 1228 root 20 0 36032 35M 1880 R 0 94.6 28.3 1088m per +l 657 root 0 0 11064 10M 1140 S 0 4.1 8.7 96:19 X 1246 scotstef 2 0 880 880 668 R 0 0.9 0.6 12:05 top

Math::Pari seems to be hung on test number 26 under t/nfields. I know this has failed a few of the graphing tests, but I don't plan (or know how) to use them.

/Pari..............FAILED tests 7, 57 + Failed 2/102 tests, 98.04% okay t/PlotRect..........# Can't locate Term/Gnuplot.pm in @INC, ignoring t +he test skipped test on this platform t/analyz............ok, 2/6 skipped: various reasons + t/elliptic..........FAILED tests 1, 4-5, 7-12, 14-15, 17-36, 38-42, 44 + Failed 37/46 tests, 19.57% okay (-3 skipped tests: 6 okay, 13.04%) t/graph.............# Can't locate Term/Gnuplot.pm in @INC, ignoring p +lotting t/graph.............FAILED tests 4, 10, 15-17, 20-21, 23 + Failed 8/28 tests, 71.43% okay (-3 skipped tests: 17 okay, 60.71%) t/linear............FAILED tests 1-2, 11, 16, 37, 75-77 + Failed 8/105 tests, 92.38% okay (-2 skipped tests: 95 okay, 90.48% +) t/nfields...........NOK 26

Is this normal? I am running this on a red hat 6.2 box with a 450mgz processor and 128 mb of ram. I also had the guy in the next cube try installing it and it got hung at the same point. Any help would be appreciated.

"The social dynamics of the net are a direct consequence of the fact that nobody has yet developed a Remote Strangulation Protocol." -- Larry Wall

Replies are listed 'Best First'.
Re: Math::Pari Taking forever and a day..
by scain (Curate) on Mar 19, 2002 at 14:01 UTC
    I don't know if it is normal, but I experienced the same thing on my RH 7.2 box. I ended up installing it by hand and killing the make test. Net::SFTP still seems to work just fine.

    Scott

(RhetTbull) Re: Math::Pari Taking forever and a day..
by RhetTbull (Curate) on Mar 19, 2002 at 17:52 UTC
    I had this same problem recently. For some reason, the version that CPAN thinks is current is actually old and tries to build with an incompatible version of libPARI. Grab the most recent version of Math::Pari from the author's site and install manually. That worked for me. The author is aware of the problem with CPAN but unfortunately powerless to fix it at the moment.
Re: Math::Pari Taking forever and a day..
by toadi (Chaplain) on Mar 19, 2002 at 15:34 UTC
    Well . Tried to install it on one gigahertz with 256mb ram laptop with RH 7.2 as OS. After two hours I quited the test and installed it without testing it further :)

    --
    My opinions may have changed,
    but not the fact that I am right