Spack安装错误

Question

我正试图通过以下教程在HPC上安装Spack：

安装Spack

我可以从GitHub克隆它，但是当我试图通过输入加载环境时：

. share/spack/setup-env.sh

我得到了下面的错误(不好意思，格式很差)。我也尝试过安装hdf5，但也出现了类似的错误。我是不是缺少了什么东西，还是应该联系维护集群的技术人员？谢谢你的帮助

Spack requires Python 2.7 or 3.5 or higher You are running spack with Python 3.4.10.

在那里，我尝试从以下位置更改setup-env.sh中的第333行：

for cmd in "${SPACK_PYTHON:-}" python3 python python2; do 

至

for cmd in "${SPACK_PYTHON:-}" python2; do

这会将python的版本更改为正确的版本，以便当我这样做时：

echo $SPACK_PYTHON
     /usr/bin/python2
/usr/bin/python2 --version
     Python 2.7.18

最后，我尝试继续本教程的输入：

spack install lib

我得到了以下错误：

​

spack install zlib
  ==
  > Bootstrapping clingo from pre - built binaries ==
  > Fetching https: //mirror.spack.io/bootstrap/github-actions/v0.2/build_cache/linux-rhel5-x86_64-gcc-9.3.0-clingo-bootstrap-spack-gypv5loj2ml73duq6sr76yg5rj25te2m.spec.json
  ==
  > Fetching https: //mirror.spack.io/bootstrap/github-actions/v0.2/build_cache/linux-rhel5-x86_64/gcc-9.3.0/clingo-bootstrap-spack/linux-rhel5-x86_64-gcc-9.3.0-clingo-bootstrap-spack-gypv5loj2ml73duq6sr76yg5rj25te2m.spack
  ==
  > Installing "clingo-bootstrap@spack%gcc@9.3.0~docs~ipo+python+static_libstdcpp build_type=Release arch=linux-rhel5-x86_64"
from a buildcache
  ==
  > Bootstrapping patchelf from pre - built binaries ==
  > Bootstrapping patchelf @0 .13 .1: 0.13 .99 % gcc target = x86_64 from sources ==
  > Installing patchelf - 0.13 .1 - ryv6d2pumi2fhaqoetncdofvtkliz2ry ==
  > No binary
for patchelf - 0.13 .1 - ryv6d2pumi2fhaqoetncdofvtkliz2ry found: installing from source ==
  > Using cached archive: /ddn/home
1 / r3031 / software / spack /
  var / spack / cache / _source - cache / archive / 08 / 08 c0237e89be74d61ddf8f6ff218439cdd62af572d568fb38913b53e222831de.tar.gz ==
  > No patches needed
for patchelf
  ==
  > patchelf: Executing phase: 'autoreconf' ==
  > patchelf: Executing phase: 'configure' ==
  > patchelf: Executing phase: 'build' ==
  > Error: ProcessError: Command exited with status 2:
  'make'
'-j16'
'V=1'

4 errors found in build log:
  36 Making all in src
37 make[1]: Entering directory '/tmp/r3031/spack-stage/spack-stage-patchelf-0.13.1-ryv6d2pumi2fhaqoetncdofvtkliz2ry/spack-src/src'
38 source = 'patchelf.cc'
object = 'patchelf.o'
libtool = no\
39 DEPDIR = .deps depmode = none / bin / sh.. / build - aux / depcomp\
40 g++ - DPACKAGE_NAME = \"patchelf\" -DPACKAGE_TARNAME=\"patchelf\" -DPACKAGE_VERSION=\"0.13.1.20211127.72b6d44\" -DPACKAGE_STRING=\"patchelf\ 0.13.1.202
11127.72 b6d44\ " -DPACKAGE_BUGREPORT=\"\" -DPACKAGE_URL=\"\" -DPACKAGE=\"patchelf\" -DVERSION=\"0.13.1.20211127.72b6d44\" -I.    -Wall -std=c++11 -D_
FILE_OFFSET_BITS = 64 - c - o patchelf.o patchelf.cc
41[spack cc] ERROR: Compiler 'gcc@4.8'
does not have a C++compiler configured. >>
  42 Makefile: 366: recipe
for target 'patchelf.o'
failed
  >>
  43 make[1]: ** * [patchelf.o] Error 1
44 make[1]: Leaving directory '/tmp/r3031/spack-stage/spack-stage-patchelf-0.13.1-ryv6d2pumi2fhaqoetncdofvtkliz2ry/spack-src/src' >>
  45 Makefile: 439: recipe
for target 'all-recursive'
failed
  >>
  46 make: ** * [all - recursive] Error 1

See build log
for details:
  /tmp/r
3031 / spack - stage / spack - stage - patchelf - 0.13 .1 - ryv6d2pumi2fhaqoetncdofvtkliz2ry / spack - build - out.txt

  ==
  > Bootstrapping clingo from sources ==
  > Error: cannot bootstrap the "clingo"
Python module from spec "clingo-bootstrap@spack+python %gcc target=x86_64"
due to the following failures:
  github - actions - v0 .2 raised RuntimeError: cannot bootstrap any of the patchelf executables from spec "patchelf@0.13.1:0.13.99 %gcc target=x86_64"
due to the following failures:
  github - actions - v0 .2 raised IOError: [Errno 2] No such file or directory: '/ddn/home1/r3031/software/spack/share/spack/bootstrap/github-actions-v0.2/patchelf.json'
github - actions - v0 .1 raised ValueError: source is not trusted
spack - install raised ChildError: ProcessError: Command exited with status 2:
  'make'
'-j16'
'V=1'
Run `spack --debug ...`
for more detailed errors
github - actions - v0 .1 raised ValueError: source is not trusted
spack - install raised ConflictsInSpecError: Conflicts in concretized spec "clingo-bootstrap@spack%gcc@4.8~docs~ipo+python~static_libstdcpp build_type=Release arch=linux-sle_hpc12-x86_64/brcc7l3"

List of matching conflicts
for spec:

  clingo - bootstrap @spack % gcc @4 .8~docs~ipo + python~static_libstdcpp build_type = Release arch = linux - sle_hpc12 - x86_64 ^
  bison @2 .7 % gcc @4 .8 arch = linux - sle_hpc12 - x86_64 ^
  cmake @3 .23 .1 % gcc @4 .8~doc + ncurses + ownlibs~qt build_type = Release arch = linux - sle_hpc12 - x86_64 ^
  ncurses @6 .2 % gcc @4 .8~symlinks + termlib abi = none arch = linux - sle_hpc12 - x86_64 ^
  pkgconf @1 .8 .0 % gcc @4 .8 arch = linux - sle_hpc12 - x86_64 ^
  openssl @1 .1 .1 o % gcc @4 .8~docs~shared certs = system arch = linux - sle_hpc12 - x86_64 ^
  perl @5 .34 .1 % gcc @4 .8 + cpanm + shared + threads arch = linux - sle_hpc12 - x86_64 ^
  berkeley - db @18 .1 .40 % gcc @4 .8 + cxx~docs + stl patches = b231fcc arch = linux - sle_hpc12 - x86_64 ^
  bzip2 @1 .0 .8 % gcc @4 .8~debug~pic + shared arch = linux - sle_hpc12 - x86_64 ^
  diffutils @3 .8 % gcc @4 .8 arch = linux - sle_hpc12 - x86_64 ^
  libiconv @1 .16 % gcc @4 .8 libs = shared, static arch = linux - sle_hpc12 - x86_64 ^
  gdbm @1 .19 % gcc @4 .8 arch = linux - sle_hpc12 - x86_64 ^
  readline @8 .1 % gcc @4 .8 arch = linux - sle_hpc12 - x86_64 ^
  zlib @1 .2 .12 % gcc @4 .8 + optimize + pic + shared patches = 0 d38234 arch = linux - sle_hpc12 - x86_64 ^
  python @2 .7 % gcc @4 .8 + bz2 + ctypes + dbm~debug + ensurepip + libxml2~nis~optimizations + pic + pyexpat + pythoncmd + readline + shared + sqlite3 + ssl~tix~tkinter + ucs4 + uuid + zlib arch = linux - sle_hpc12 - x86_64 ^
  re2c @2 .2 % gcc @4 .8 arch = linux - sle_hpc12 - x86_64

1. "%gcc@:5"
conflicts with "clingo-bootstrap" [C++14 support is required to bootstrap clingo]

Run `spack --debug ...`
for more detailed errors

​

可能有一些有用的信息:在/usr/bin/中，我有以下版本的python：

​

​

Answer 1

我看不出有什么问题，所以试着留下一些提示。

第一个问题是，您不需要破解shell脚本来强制使用Python版本。你可以简单地：

export SPACK_PYTHON=<python-you-want-to-use>

例如，您可以指向Spack来使用您似乎在机器上拥有的Python3.6解释器。

第二个技巧是您可以使用以下Docker映像：

$ docker pull ghcr.io/spack/tutorial:latest

要阅读本教程，例如您自己的笔记本电脑，而不是您可以访问的HPC集群。这将给您在集群上使用Spack之前熟悉它的时间。

希望这能有所帮助。