A toolkit for Grid Computing. I found a nice article at http://www.linux-magazin.de/Artikel/ausgabe/2004/06/globus/globus.html about it. Very nice.
I already played around with globus and i have a first ebuild for it. It still has some problems with the build system from globus... I will attach it later, so anyone who wants can fix it. Once it is working i will check it in.
Created attachment 31574 [details] globus-toolkit-3.0.2.ebuild This ebuild was for testing only first. When i call the globus installer by hand everything works as it should. When i call it from inside an ebuild it stops with an error message. I tried to figure out why and i think it's ccache related or it uses some variables that are already used by portage.
Any news on this? Did anyone play around with the ebuild i submitted?
How do I get rid of the message: "[empty/missing/bad digest]" ?
change to the dir where the ebuild is located and execute an ebuild globus-toolkit-3.0.2.ebuild digest it will fetch the sources needed and compute the digest. afterwards you should be able to emerge the ebuild with ebuild globus-toolkit-3.0.2.ebuild merge or by using emerge.
Rene: any news on this bug?
Created attachment 39720 [details] Ebuild for globus 3.2.1 Updated ebuild for globus 3.2.1. Still fails.
Hi I ported the ebuild to Globus Toolkit 3.2.1 (see attachement) and removed the ccache stuff. I get the following error when emerging it: /usr/bin/gcc -DHAVE_CONFIG_H -I. -I/mnt/rwstorage/var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-insta ller/BUILD/globus_common-4.10/library -I.. -I. -I/var/tmp/portage/globus-toolkit-3.2.1/image/opt/globus-toolkit-3.2. 1/include/gcc32dbgpthr -I/var/tmp/portage/globus-toolkit-3.2.1/image/opt/globus-toolkit-3.2.1/include -I/var/tmp/por tage/globus-toolkit-3.2.1/image/opt/globus-toolkit-3.2.1/include/gcc32dbgpthr -g -march=pentium3 -O2 -pipe -D_REENTR ANT -Wall -c globus_debug.c -fPIC -DPIC -o .libs/globus_debug.lo globus_debug.c:7: error: parse error before "__extension__" globus_debug.c:7: error: `__len' undeclared here (not in a function) globus_debug.c:7: error: initializer element is not constant globus_debug.c:7: error: parse error before "if" globus_debug.c:7: warning: type defaults to `int' in declaration of `__retval' globus_debug.c:7: error: conflicting types for '__retval' globus_debug.c:7: error: previous definition of '__retval' was here globus_debug.c:7: warning: data definition has no type or storage class globus_debug.c:7: error: parse error before '}' token make[2]: *** [globus_debug.lo] Error 1 make[2]: Leaving directory `/mnt/rwstorage/var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BU ILD/globus_common-4.10/library' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/mnt/rwstorage/var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BU ILD/globus_common-4.10' make: *** [all-recursive-am] Error 2 But when running install-gt3 manually with the same arguments, everything works fine and dandy, strange...
Globus is turning into a monstrosity. Therefore, I really think we might be better off having GPT as a seperate ebuild, and then having globus split into different ebuilds too. Also, I got exactly the same problem trying out this ebuild.
Created attachment 49555 [details] Working ebuild for gt-3.2.1 This is a working ebuild for the Globus middleware version 3.2.1 with the security advisories. Tested on x86 and ppc (pegasos), should work on any 32 bit architecture. For 64 bit architectures a small change only should be needed (will provide later)
Created attachment 49556 [details, diff] patch required for the ebuild A few changes required in the Globus build system
Created attachment 50063 [details] new version Updated version - should theoretically work on the 64 bit architectures and a small (but quite important) typo fixed. I also bundled gsi_ssh and myproxy packages - I tried to make a separate ebuilds for them (and maybe later create a split globus ebuid based on the experience gained this way) but gpt (the Globus packaging system) is not very cooperative - it wants to have all the previously installed files available and install into the same directory tree, which won't work with sandbox certainly... There's place for 2 USE flags (for ssh and myproxy), but I left it out for now. The previously supplied patch is still necessary
Can you please give me a hint which jdk version you used? with sun 1.4.2.06 it fails: BUILD FAILED java.lang.UnsupportedClassVersionError: jdepend/xmlui/JDepend (Unsupported major.minor version 49.0) with sun 1.5.0.01 it fails with errors regarding enums: [javac] (try -source 1.4 or lower to use 'enum' as an identifier)
on the x86: java -version java version "1.4.2_06" Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_06-b03) Java HotSpot(TM) Client VM (build 1.4.2_06-b03, mixed mode) on ppc: java -version java version "1.3.1" Java(TM) 2 Runtime Environment, Standard Edition (build Blackdown-1.3.1-02c-FCS) Classic VM (build Blackdown-1.3.1-02c-FCS, native threads, nojit) Which actually should fail - Globus team recommends version >=1.4.2 - I'm testing with IBM 1.4.2 right now.
Could you test with dev-java/sun-jdk Latest version available: 1.4.2.06 and report back if it works for you? Over here it does not work... nor does sun 1.5....
We're talking x86 here? If so: emerge -pv dev-java/sun-jdk [ebuild Rf ] dev-java/sun-jdk-1.4.2.06 -doc -gnome -jce -kde -mozilla 0 kB It's the version I'm using right now. I suppose the error comes from some other package (part of ant distro perhaps?) 'cause at this stage the jdk is used in a very basic way - just to run ant I think.
mobile globus-toolkit # emerge globus-toolkit-3.2.1.ebuild emerging by path implies --oneshot... adding --oneshot to options. *** emerging by path is broken and may not always work!!! Calculating dependencies ...done! >>> emerge (1 of 1) sys-cluster/globus-toolkit-3.2.1 to / >>> md5 src_uri ;-) gt3.2.1-all-source-installer.tar.gz >>> md5 src_uri ;-) globus_xio-0.11.tar.gz >>> md5 src_uri ;-) globus_io-5.5.tar.gz >>> md5 src_uri ;-) globus_nexus-6.5.tar.gz >>> md5 src_uri ;-) globus_ftp_client-2.9.tar.gz >>> md5 src_uri ;-) globus_gram_client_tools-4.3.tar.gz >>> md5 src_uri ;-) gsi_openssh_bundle-3.5-src.tar.gz >>> md5 src_uri ;-) myproxy-1.16.tar.gz >>> Unpacking source... * Applying globus-fix.diff ... [ ok ] >>> Source unpacked. >>> Test phase [not enabled]: sys-cluster/globus-toolkit-3.2.1 >>> Install globus-toolkit-3.2.1 into /var/tmp/portage/globus-toolkit-3.2.1/image/ category sys-cluster build_gpt ====> installing GPT into /var/tmp/portage/globus-toolkit-3.2.1/image//usr/local/globus build_gpt ====> building support/Compress-Zlib-1.16 build_gpt ====> building support/Archive-Tar-0.22 build_gpt ====> building support/PodParser-1.18 build_gpt ====> building support/Digest-MD5-2.20 build_gpt ====> building packaging_tools Build environment: ant is /usr/bin/ant java is /opt/sun-jdk-1.4.2.06/bin/java gcc is /usr/lib/ccache/bin/gcc gpt-build ====> Changing to /var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BUILD/globus_core-3.14/ gpt-build ====> BUILDING FLAVOR gcc32dbg gpt-build ====> Changing to /var/tmp/portage/globus-toolkit-3.2.1/image/usr/local/globus/etc gpt-build ====> REMOVING empty package globus_core-gcc32dbg-pgm_static gpt-build ====> REMOVING empty package globus_core-noflavor-doc gpt-build ====> CHECKING BUILD DEPENDENCIES FOR globus_core SKIPPING REBUILD of gcc32dbg gpt-build ====> CHECKING BUILD DEPENDENCIES FOR core gpt-build ====> Changing to /var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BUILD/core-src/ gpt-build ====> BUILDING FLAVOR ant -f impl/java/build.xml dist Buildfile: impl/java/build.xml setenv: [mkdir] Created dir: /var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BUILD/core-src/impl/java/build/classes [mkdir] Created dir: /var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BUILD/core-src/impl/java/build/lib [mkdir] Created dir: /var/tmp/portage/globus-toolkit-3.2.1/work/gt3.2.1-all-source-installer/BUILD/core-src/impl/java/schema BUILD FAILED java.lang.UnsupportedClassVersionError: jdepend/xmlui/JDepend (Unsupported major.minor version 49.0) Total time: 1 second ERROR: Build has failed Error building gt3-all-src !!! ERROR: sys-cluster/globus-toolkit-3.2.1 failed. !!! Function src_install, Line 87, Exitcode 1 !!! install-gt3 failed !!! If you need support, post the topmost build error, NOT this status message.
I reemerged ant, it changed nothing, atm i have no clue what goes wrong with the ebuild. And it would be great if the ebuild depends on either sun-jdk-1.4.2 or handles 1.5.0 correctly by setting the above mentioned flag in that case.
Created attachment 50119 [details] updated jdk depend Changed the depend to =virtual/jdk-1.4* . Making globus work with jdk 1.5.0 is not a trivial task (as you probably know). I checked the ant task which is failing and it's extremely trivial: it simply copies a single file (.jar) into a directory (created as the previous message states). It cannot fail because of Globus problem - it's a very standard ant task having nothing Globus-specific in it. If you could please unemerge jdk 1.5.0 (it's hardmasked for a reason I suppose) and give it another try it might success.
I think unmerging jdk-1.5.0 is no option as i'm using it heavily for devlopement. So the error is just a wrong version of that one .jar? Couldn't the ebuild take care of that in some way?
I'm not much of a java developer, so I might be wrong, but I don't see a way to take care of an incorrect jar version (except messing with ants CLASSPATH), especially that I don't know which jar it might be and where could the proper version be - as I said it fails on basic ant functionality not on some compillation or Globus dependant part. There's the section of build.xml that gives the error: <mkdir dir="${schema.dir}"/> <copy toDir="${schema.dir}"> <fileset dir="${schema.origin}" casesensitive="yes"/> </copy> <copy file="${nsmap.file}" toDir="${schema.dir}"/> <copy file="core-config.wsdd" toFile="server-config.wsdd"/> <copy file="core-bundles.properties" toFile="${bundle.file}"/> <copy file="core-client-config.wsdd" toFile="client-config.wsdd"/> <copy file="core-ogsilogging.properties" toFile="ogsilogging.properties"/> <available property="post-compile.present" file="post-compile.xml" /> <available property="stubs.present" type="dir" file="${stubs.dir}" /> <available property="schema.present" type="file" file="${schema.dir}/ogsi/ogsi.wsdl" /> <available property="key.present" file="${keystore}" /> <mkdir dir="${stubs.dir}"/> The first mkdir gets executed and the last doesn't. I don't see anything even remotely suspicious here so I sustain my opinion that something is wrong with your ant and it might be connected with jdk 1.5.0 being around. Please try running the globus install by hand to determine whether it's problem with the ebuild or something deeper.
I am going to provide a newer ebuild for v4.0 has anyone done any work on one yet?
Hanni, I was about to start, but I'd love to see your 4.0 attempt.
Hi, I've made a start here's the ebuild it just configures and compiles, no install. without: sed -e "s|GLOBUS_LOCATION=@prefix@|GLOBUS_LOCATION=\${D}@prefix@| ; \ s|GLOBUS_IODBC_PATH=@globus_iodbc@|GLOBUS_IODBC_PATH=\${D}@globus_iodbc@|" \ Makefile.in-orig > Makefile.in There are loads of ACCESS DENIED errors. If you add --with-iodbc=/usr/lib to ./configure it fails with /usr/lib/ccache/bin/gcc -DPACKAGE_NAME=\"\" -DPACKAGE_TARNAME=\"\" -DPACKAGE_VERSION=\"\" -DPACKAGE_STRING=\"\" -DPACKAGE_BUGREPORT=\"\" -DPACKAGE=\"globus_rls_client\" -DVERSION=\" 4.1\" -I. -I/var/tmp/portage/gt-4.0.2/work/gt4.0.2-all-source-installer/source-trees-thr/replica/rls/client ./library -I/var/tmp/portage/gt-4.0.2/image//usr/local/globus-4.0.2/include -I/var/tmp/portage/gt-4.0.2/image//usr/local/globus- 4.0.2/include/gcc32dbgpthr -g -O2 -march=pentium4 -fomit-frame-pointer -D_REENTRANT -Wall -c cli.c gcc: ./library: linker input file unused because linking not done gcc: ./library: linker input file unused because linking not done cc1: ./library: No such file or directory make[2]: *** [cli.o] Error 1 make[2]: Leaving directory `/var/tmp/portage/gt-4.0.2/work/gt4.0.2-all-source-installer/source-trees-thr/replica/rls/client' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/var/tmp/portage/gt-4.0.2/work/gt4.0.2-all-source-installer/source-trees-thr/replica/rls/client' ERROR: Build has failed make: *** [globus_rls_client-thr] Error 2 !!! ERROR: sys-cluster/gt- 4.0.2 failed. Call stack: ebuild.sh, line 1527: Called dyn_compile ebuild.sh, line 931: Called src_compile gt-4.0.2.ebuild, line 50: Called die If you have the time I'd appreciate and help/advice. This is the first ebuild I've bothered trying so feel free to help me learn. I pulled all the dep info and other stuff from here http://www.globus.org/toolkit/docs/4.0/admin/docbook/quickstart.html#q-toolkit this ebuild currently lacks support for a Scheduler and use flags will be implemented for optional aspects like scheduler selection and IODBC. I've written this out on my desktop but will install it to my cluster probably this weekend. At some point I will check it with other java environments (this one uses SUN). I've only tried it on x86 32 bit, P4 my cluster is made up of VIA C3's and I'll have another one of VIA C7's to test with in about a month. I haven't tried running the toolkit yet but will, over the next few weeks. Hanni
Created attachment 87442 [details] gt -4.0.2 ebuild gt-4.0.2 ebuild progress, no install implemented yet
Hanni, why do you escape the ${D} in the sed statements? I wonder if this is why your install is failing? Can you paste the install error?
(In reply to comment #26) > Hanni, why do you escape the ${D} in the sed statements? I wonder if this is > why your install is failing? Can you paste the install error? > Sorry for any confusion, but you've missunderstood. The compile works fine with that ebuild, I haven't added the install part yet. If you try the ebuild _without_ the sed part you recieve a lot of access denied errors and the make fails. The escape may not be necessary actually, I'll check. When I tried to build with IODBC support it fails, I just need to investigate this a little more. Maybe something to do with differences between libiodbc gentoo package and libiodbc2 libiodbc2-dev mentioned in the guide. I'll sort out the install part this weekend. If anyone wants to check if it compiles ok on other architectures that would be cool. Hanni
Hi Hanni, sorry for the delay. I had to tweak your ebuild a slight bit (adding an additional sed expression) and it started building quite well. Tomorrow morning I'll post an updated ebuild. Thanks! Seemant
Created attachment 88631 [details] updated ebuild Hanni, try this one out.
(In reply to comment #29) > Created an attachment (id=88631) [edit] > updated ebuild > > Hanni, try this one out. > That's great although for IODBC support to actually be compiled into globus you need to add --with-iodbc=/usr/lib to .configure ./configure --with-iodbc=/usr/lib || die "./configure failed" However I feel IODBC support should be activated through a use flag, but we'll sort that out in a bit. I've not done much work on install yet have you? Hanni
omg, what a broken ass build system this globus thing has! Wow! the files in ${D}/usr/local/ get created during src_compile (you can verify) and then for some unknown reason, everything gets removed after the compile is done! What the hell are the globus people smoking to create a monstrosity like this? Damon, Konrad, you guys have any luck with 4.0.2? Damon, I'm especially interested in the separate packaging for gpt if you have any ideas about it. Talk to me people!
Created attachment 89164 [details] Perl portion of the deps The attached tar file contains makepatch and the two perl deps that aren't in the tree.
Created attachment 100185 [details] GT 4.0.3 ebuild This is an ebuild which gets the Globus Toolkit v 4.0.3 installed on two of my x86 machines. It's missing a whole bunch of deps and just barely works... But hopefully it will help some people out. (requires a one-line patch I'm also attaching)
Created attachment 100186 [details, diff] one line patch to a source file in gt4.0.3 a patch to fix a template file in wsrf-cgen... without this patch, g++ 4.1 bombs while compiling generated parsers.
I'll have a look, need globus anyway. Thanks for the work you put up so far :)
*** Bug 212175 has been marked as a duplicate of this bug. ***
Created attachment 152713 [details] updated ebuild
thank you guy! I'd like to update this 'bug' upgrading your ebuold for the latest available globus toolkit ver 4.0.7. There are not many changes to your ebuild, only a version (and mirror) update. I've succesfully installed it on and amd64 system.(In reply to comment #34) > Created an attachment (id=100186) [edit] > one line patch to a source file in gt4.0.3 thank you guy! I'd like to update this 'bug' upgrading your ebuild for the latest available globus toolkit ver 4.0.7. There are not many changes to the old ebuild, only a version (and mirror) update. I've succesfully installed it on and amd64 system. bye
Created attachment 153037 [details] Fully functional 4.0.7 + patches Hy, sorry for my first bad ebuild... this is a fully working version with all patches available applied. See http://www.globus.org/toolkit/advisories.html?version=4.0 for more information. N.B.: this ebuild need Make.patch in the sys-cluster/gt/file directory!
Created attachment 153039 [details, diff] patch for fully working gt-4.0.7.ebuild MakeFile patch to build openssh-4.3
Hi everybody Thanks for the ebuild for version 4.0.7. I have a problem with installing rls, because it cannot find some stuff from odbc. Tried it with installing myodbc too. Then it went smoothly. Not sure if a other odbc driver works too. But at least one dependency is missing. Globus seems to look directly for the driver instead of the interface. Thanks, Marc
Hi I suggesst to use dev-db/unixODBC instead of dev-db/myodbc. Hope this is the minimal solution to compile the rls stuff... Thanks, Marc
i added some dep to the ebuild: dev-db/unixODBC dev-perl/XML-Parser and use this config: ./configure --prefix=${GLOBUS} \ $(use_enable lsf wsgram-lsf) \ $(use_enable pbs wsgram-pbs) \ --disable-prewsgram \ --enable-gridway \ -with-iodbc=/usr/lib/ \ only do make to get full install of globus # cd ${WORKDIR}/gt${PV}-all-source-installer make # Yep, compile during the install phase. Fix it if you can :) make install || die "make install failed!" at the moment (i'm still configuring) i've no problem and it's works out of the box. anyone have a good init script for the ws-container?? i wrote this one: #!/sbin/runscript export GLOBUS_LOCATION=/opt/globus-4.0.7 export GLOBUS_OPTIONS="-Xms256M -Xmx512M" source $GLOBUS_LOCATION/etc/globus-user-env.sh depend() { need net } start() { ebegin "Starting Globus WS Container" ${GLOBUS_LOCATION}/bin/globus-start-container ${CONTAINER_OPTS} > ${LOG_FILE} 2>&1 & eend ${?} } stop() { ebegin "Stopping Globus WS Container" ${GLOBUS_LOCATION}/bin/grid-proxy-init \ -cert /etc/grid-security/containercert.pem \ -key /etc/grid-security/containerkey.pem \ -out /tmp/containerproxy.pem > /dev/null export X509_USER_PROXY=/tmp/containerproxy.pem ${GLOBUS_LOCATION}/bin/globus-stop-container export X509_USER_PROXY rm /tmp/containerproxy.pem eend ${?} } it's not the best way, but it works for me!! thanks to all for your work!!
(this is an automated message based on filtering criteria that matched this bug) Hello, The Gentoo Team would like to firstly thank you for your ebuild submission. We also apologize for not being able to accommodate you in a timely manner. There are simply too many new packages. Allow me to use this opportunity to introduce you to Gentoo Sunrise. The sunrise overlay[1] is a overlay for Gentoo which we allow trusted users to commit to and all users can have ebuilds reviewed by Gentoo devs for entry into the overlay. So, the sunrise team is suggesting that you look into this and submit your ebuild to the overlay where even *you* can commit to. =) Because this is a mass message, we are also asking you to be patient with us. We anticipate a large number of requests in a short time. Thanks, On behalf of the Gentoo Sunrise Team, Jeremy. [1]: http://www.gentoo.org/proj/en/sunrise/ [2]: http://overlays.gentoo.org/proj/sunrise/wiki/SunriseFaq
Created attachment 201255 [details] updated ebuild for globus-4.2.1
Created attachment 201256 [details] env.d script
Created attachment 201257 [details] gaa_test patch
Created attachment 201258 [details] init.d script
Created attachment 201260 [details] patch to open in gwd which caused compilation error
Created attachment 201262 [details] globus metadata
Created attachment 201264 [details] seperate ebuild for the gpt build tools
Created attachment 201266 [details] env.d entry
Created attachment 201267 [details] globus-build metadata
attached are two updated ebuilds for globus-build and globus. Globus-build builds the gpt tools separately as requested. The globus ebuild builds all the standard targets, and is set up to also do all the targets mentioned in the INSTALL document (eg. i18n). Both of these build into a scratch directory in ${S}/build so as to not build directly into the install tree. These ebuilds have passed the Sunrise review, but was requested to do a further Java review. I am posting these here in the hopes that they are useful, and that they can be further reviewed by people using/maintaining globus. Best regards, EBo --
(In reply to comment #54) > attached are two updated ebuilds for globus-build and globus. Globus-build > builds the gpt tools separately as requested. The globus ebuild builds all the > standard targets, and is set up to also do all the targets mentioned in the > INSTALL document (eg. i18n). Both of these build into a scratch directory in > ${S}/build so as to not build directly into the install tree. > > These ebuilds have passed the Sunrise review, but was requested to do a further > Java review. I am posting these here in the hopes that they are useful, and > that they can be further reviewed by people using/maintaining globus. > > Best regards, > > EBo -- > what further steps are necessary to setup globus? I have emerged the ebuilds and followed the first steps of the official SysAdmin Guide, setup a simple CA etc. however, when starting globus I get java exceptions like "Error: ; nested exception is: java.net.ConnectException: Connection refused" I don't know where to dig now, to resolve this issue...
Johannes, The ebuilds were released to early -- I was just returning to school and knew that if I did not post what I had it would be many months later. I still have not resolved with the universities lawyers as to whether I can work on the local super-computers or not, and the ports are being blocked for security reasons. I've gotten about as far as I can on my laptop.... Once I sort all that out I can get back to testing. I expect it to take a couple of weeks more. Please email me back at if I have not readdress the issue by then. Sorry for the delay. EBo --
(In reply to comment #56) > Johannes, > > The ebuilds were released to early -- I was just returning to school and knew > that if I did not post what I had it would be many months later. I still have > not resolved with the universities lawyers as to whether I can work on the > local super-computers or not, and the ports are being blocked for security > reasons. I've gotten about as far as I can on my laptop.... Once I sort all > that out I can get back to testing. I expect it to take a couple of weeks > more. > > Please email me back at if I have not readdress the issue by then. > > Sorry for the delay. > > EBo -- > hello EBo, you have done a great job. I could solve the issues I had. I had errors about an "ambiguous redirect" in the init script, which I could resolve. There have been some problems concering the environmental variables, which I could fix now either. I can now submit simple jobs. It's getting complicated now, because the documentation I have found on working with and testing services/clients is either outdated or things differ a lot from my installation. greetings, Johannes
Johannes, Thanks, it is really rough, but maybe between all our fixes we can get things all sorted out. I'll try to take a look at this in a week or two after I get back from a business trip... Thanks!
Anyone interested in pre-maintainship in the science overlay?
(In reply to comment #59) > Anyone interested in pre-maintainship in the science overlay? Sébastien, I'm not sure what you mean by pre-maintainship, but I did install my experimental version in the science-overlay. BTW, sorry for disappearing -- I caught a NASTY cold after returning abroad, and have just now *almost* caught up with with my obligations. Maybe if someone can contact me off-list we can bang it better into shape. EBo --
I can't find this in the science overlay. Is it under a different name?
I had a version uploaded onto the sunrise overlay and later decided to have it pulled. I decided to use Plan9 for my research instead of the MPI and Globus tools. I had not gotten Globus to the point I thought it was dependable due to its byzantinen build and configuration issues.
Is there someone still interested in this project and having it in the tree? Version 6.0 is out. http://toolkit.globus.org/ftppub/gt6/installers/src/globus_toolkit-6.0.1433516164.tar.gz I just downloaded the sources, extracted them and ran the ./configure script. It's been over 10 minutes and counting now... What the eff?! (In reply to John (EBo) David from comment #62) > I had not gotten Globus to the point I thought it was > dependable due to its byzantinen build and configuration issues. I don't know how long it was taking before to configure and build globus but it's definitely got worse over time. Please comment or I will close this bug as WONTFIX.
It has been SO LONG since I looked at this that it is years beyond my even caring. I scrapped using Globus and started playing with 9P, and then scrapped that... If my vote counts any, mark this bug as WONTFIX, WTFIGO, or whatever else is appropriate. Thanks for checking back into this though...
Thanks John for your input. As for for the others following this bug, I'm going to go ahead and mark this bug as WONTFIX since nobody commented and there hasn't been much activity since 2010. Feel free to reopen/revisit this bug.