ml-it-r-devel at epigenomics.com
2009-Apr-16 15:37 UTC
[Rd] R (2.9.0 rc) CMD INSTALL will leave OOLOCK folder even if installation succeeded
Hi, since a couple of days ago I see failures when using R CMD INSTALL -l lib path-to-src/pkg.tar.gz This occurs in automated library updating via a shell script and is caused by the immediate preceding call to leave an 00LOCK folder even though the package installation of that call succeeded. This occurs at random (no obvious/systematic correlation to package name or or such) and only for packages where an earlier version has been installed into the same library location. e.g. ~/R/hardy$ R-2.9.0/bin/R CMD INSTALL -l R-2.9.x-libs R-2.9.x-libsSrc/CRAN/feature_1.2.3.tar.gz ... ** building package indices ... * DONE (feature) but ~/R/hardy$ ls -al R-2.9.x-libs/00LOCK/feature/ total 0 lrwxrwxrwx 1 XXX epiR 39 Mar 27 05:56 feature -> /mnt/local/R/hardy/R-2.9.x-libs/feature/ ?INSTALL tells me of a workaround via --unsafe or --no-lock but I would prefer to use the default incantation. Any insights what the cause could be? Regards, Matthias R version 2.9.0 RC (2009-04-10 r48321) on Ubuntu hardy -- Matthias Burger Project Manager/ Biostatistician Epigenomics AG Kleine Praesidentenstr. 1 10178 Berlin, Germany phone:+49-30-24345-0 fax:+49-30-24345-555 http://www.epigenomics.com matthias.burger at epigenomics.com -- Epigenomics AG Berlin Amtsgericht Charlottenburg HRB 75861 Vorstand: Geert Nygaard (CEO/Vorsitzender) Oliver Schacht PhD (CFO) Aufsichtsrat: Prof. Dr. Dr. hc. Rolf Krebs (Chairman/Vorsitzender)
Peter Dalgaard
2009-Apr-16 16:25 UTC
[Rd] R (2.9.0 rc) CMD INSTALL will leave OOLOCK folder even if installation succeeded
ml-it-r-devel at epigenomics.com wrote:> Hi, > > since a couple of days ago I see failures when using > R CMD INSTALL -l lib path-to-src/pkg.tar.gz > > This occurs in automated library updating via a shell script and is caused by the > immediate preceding call to leave an 00LOCK folder > even though the package installation of that call succeeded. > This occurs at random (no obvious/systematic correlation to package name or or such) and > only for packages where an earlier version has been installed into the same library location. > > e.g. > > ~/R/hardy$ R-2.9.0/bin/R CMD INSTALL -l R-2.9.x-libs R-2.9.x-libsSrc/CRAN/feature_1.2.3.tar.gz > > ... > ** building package indices ... > * DONE (feature) > > but > ~/R/hardy$ ls -al R-2.9.x-libs/00LOCK/feature/ > total 0 > lrwxrwxrwx 1 XXX epiR 39 Mar 27 05:56 feature -> /mnt/local/R/hardy/R-2.9.x-libs/feature/ > > ?INSTALL tells me of a workaround via --unsafe or --no-lock but I would prefer to use the > default incantation. > Any insights what the cause could be? > > Regards, Matthias > > > R version 2.9.0 RC (2009-04-10 r48321) > on Ubuntu hardy >Not (easily) reproducible for me on r48325 (which only differs from 48321 in the FAQ and package count in internet.Rout.save). So, cannot possibly fix for 2.9.0 tomorrow. Anything peculiar about your setup (e.g., networked drive) that might keep the unlink from happening? -- O__ ---- Peter Dalgaard ?ster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907
Possibly Parallel Threads
- R 2.0.0: problem: installing --with-package-versions
- installing package with version number using namespaces & dynamic library
- R 2.0.0: namespaces, S4 classes & versioned package installation: failure to resolve correct pkg version
- R 2.9.0 devel: package installation with configure-args option
- ifelse behaviour