trouble installing packages in CentOS: internet routines cannot be loaded

Question

I am having some trouble installing packages in R on CentOS after updating R. Here is an example of what happens when I try to install a package in R:

> install.packages("ggplot2")
Installing package into '/home/albers/R/x86_64-pc-linux-gnu-library/3.2'
(as 'lib' is unspecified)
--- Please select a CRAN mirror for use in this session ---
Error in download.file(url, destfile = f, quiet = TRUE) :
  internet routines cannot be loaded
In addition: Warning message:
In download.file(url, destfile = f, quiet = TRUE) :
  unable to load shared object '/usr/local/lib64/R/modules//internet.so':
  /usr/local/lib64/R/modules//internet.so: undefined symbol: curl_multi_wait
HTTPS CRAN mirror

 1: 0-Cloud [https]                2: Austria [https]
 3: Chile [https]                  4: China (Beijing 4) [https]
 5: Colombia (Cali) [https]        6: France (Lyon 2) [https]
 7: France (Paris 2) [https]       8: Germany (M▒nster) [https]
 9: Iceland [https]               10: Mexico (Mexico City) [https]
11: Russia (Moscow) [https]       12: Spain (A Coru▒a) [https]
13: Switzerland [https]           14: UK (Bristol) [https]
15: UK (Cambridge) [https]        16: USA (CA 1) [https]
17: USA (KS) [https]              18: USA (MI 1) [https]
19: USA (TN) [https]              20: USA (TX) [https]
21: USA (WA) [https]              22: (HTTP mirrors)


Selection: 1
Warning: unable to access index for repository https://cran.rstudio.com/src/contrib:
  internet routines cannot be loaded
Warning message:
package 'ggplot2' is not available (for R version 3.2.5)

So I spent some time searching the error messages. I updated my version of curl which now is up to date:

$ curl -V
curl 7.48.0 (x86_64-pc-linux-gnu) libcurl/7.48.0 zlib/1.2.8
Protocols: dict file ftp gopher http imap pop3 rtsp smtp telnet tftp
Features: IPv6 Largefile libz UnixSockets

Still unable to install packages and am receiving the same error message. So I tried following the instructions of this post , specifically adding the following command to my .bash_profile:

export LD_LIBRARY_PATH=$TOOLS/curl-7.48.0/lib:$LD_LIBRARY_PATH

Again I get the same error message about R unable to load shared objects. Ultimately this is a download issue as download.file fails as well. Here is my sessionInfo():

R version 3.2.5 (2016-04-14)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS release 6.2 (Final)

locale:
 [1] LC_CTYPE=en_US.iso885915       LC_NUMERIC=C
 [3] LC_TIME=en_US.iso885915        LC_COLLATE=en_US.iso885915
 [5] LC_MONETARY=en_US.iso885915    LC_MESSAGES=en_US.iso885915
 [7] LC_PAPER=en_US.iso885915       LC_NAME=C
 [9] LC_ADDRESS=C                   LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

I am able to download files with wget and curl but as soon as I start R, I unable to download anything.

Does anyone have any suggestions on how I might get around this issue?

Kizer

Answer 1

Try it like this.

install.packages("ggplot2", repos="http://cran.cnr.berkeley.edu")

or just before you install it change your download tool to something else ie on Linux you can try wget as follows.

options(download.file.method = "wget")

Note, I've dropped the https for http. I had a similar error...

Selection: 1
Warning: unable to access index for repository https://cran.rstudio.com/src/contrib:
  unsupported URL scheme
Warning message:
package ‘ggplot2’ is not available (for R version 3.2.3) 

If you look at the output of the following line from your question...

Yours
Protocols: dict file ftp gopher http imap pop3 rtsp smtp telnet tftp
Mine (Also Centos 6)
Protocols: tftp ftp telnet dict ldap ldaps http file https ftps scp sftp 

Your curl is missing https support. There was an error regarding curl_multi_wait in your output as well. This was added in 7.28.0.

https://curl.haxx.se/libcurl/c/curl_multi_wait.html

If you need to build curl locally do this

mkdir -p $HOME/curl
cd $HOME/curl
git clone https://github.com/curl/curl.git
cd curl
bash buildconf
./configure --prefix=$HOME/libcurl
make
make install

At this point you need to setup your LD_LIBRARY_PATH in .bash_profile ie

export LD_LIBRARY_PATH=$HOME/libcurl/lib:$LD_LIBRARY_PATH

Then either logout and back in again or

source ~/.bash_profile

Sargent

Answer 2

I believe I answered a similar question once before following the R 3.2.* upgrade last summer. Travis CI using Ubuntu 12.04 has similar issues.

One rather simple way around it is to tell R to use the external wget or curl binary you may have (and which needs to be https-ready) to download. So try eg

R> options("download.file.method"="wget")

which, when set, allows me to download your desired package from the desired non-https host:

R> install.packages("ggplot2", repos="http://cran.cnr.berkeley.edu")
Installing package into ‘/usr/local/lib/R/site-library’
(as ‘lib’ is unspecified)
--2016-04-26 18:29:16--  http://cran.cnr.berkeley.edu/src/contrib/ggplot2_2.1.0.tar.gz
Resolving cran.cnr.berkeley.edu (cran.cnr.berkeley.edu)... 169.229.201.201, 2607:f140:0:8000::201
Connecting to cran.cnr.berkeley.edu (cran.cnr.berkeley.edu)|169.229.201.201|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 1571788 (1.5M) [application/x-gzip]
Saving to: ‘/tmp/RtmpwWvA3s/downloaded_packages/ggplot2_2.1.0.tar.gz’

/tmp/RtmpwWvA3s/downloaded_packages/ggplot2_2.1.0.ta 100%[========================================================================================================================>]   1.50M  1.32MB/s   in 1.1s   

2016-04-26 18:29:17 (1.32 MB/s) - ‘/tmp/RtmpwWvA3s/downloaded_packages/ggplot2_2.1.0.tar.gz’ saved [1571788/1571788]

* installing *source* package ‘ggplot2’ ...
** package ‘ggplot2’ successfully unpacked and MD5 sums checked
** R
** data
*** moving datasets to lazyload DB
** inst
** preparing package for lazy loading
** help
*** installing help indices
** building package indices
** installing vignettes
** testing if installed package can be loaded
* DONE (ggplot2)

The downloaded source packages are in
        ‘/tmp/RtmpwWvA3s/downloaded_packages’

You can set options in ~/.Rprofile or the corresponding Rprofile.site.

Recurvate