[R] Very odd issue getting "Input/output error" running shiny app from index.html locally but same shiny on my server working fine

Chris Evans chr|@ho|d @end|ng |rom p@yctc@org
Mon May 19 17:51:01 CEST 2025


Oooh thanks.  I should probably have thought of that: this list is 
brilliant when my brain goes on strike.

OK, it's http://127.0.0.1:5071 for me and using the "view the page with 
the system web browser" when I have the index.html page opened via 
Rstudio has the system browser point to 
file:///run/user/1000/doc/9065bcbf/index.html and if I click on the link 
to the app it goes to 
file:///run/user/1000/doc/9065bcbf/apps/COREpapers1 and says "Your file 
couldn’t be accessed"

Then if I open a console with root privileges I get to this:
     root using tuxedo-os:/run/user/1000# cd doc
     bash: cd: doc: Permission denied
     root using tuxedo-os:/run/user/1000# ls doc
     ls: cannot access 'doc': Permission denied

and this:

    root using tuxedo-os:/run/user# ls -lsart
    total 0
    0 drwxr-xr-x  3 root  root    60 May 19 15:36 .
    0 drwx------ 21 chris chris  660 May 19 16:42 1000
    0 drwxr-xr-x 43 root  root  1160 May 19 16:46 ..

and


    root using tuxedo-os:/run/user/1000# ls -lsart
    ls: cannot access 'doc': Permission denied
    ls: cannot access 'gvfs': Permission denied
    total 12
    ? d?????????  ? ?     ?       ?            ? gvfs
    ? d?????????  ? ?     ?       ?            ? doc
    0 srw-rw-rw-  1 chris chris   0 May 19 15:36 bus
    0 drwx------  2 chris chris  60 May 19 15:36 gcr
    0 srw-------  1 chris chris   0 May 19 15:36 drkonqi-coredump-launcher
    0 drwxr-xr-x  2 chris chris  60 May 19 15:36 speech-dispatcher
    0 srw-rw-rw-  1 chris chris   0 May 19 15:36 snapd-session-agent.socket
    0 srw-rw-rw-  1 chris chris   0 May 19 15:36 pk-debconf-socket
    0 srw-rw-rw-  1 chris chris   0 May 19 15:36 pipewire-0-manager
    0 srw-rw-rw-  1 chris chris   0 May 19 15:36 pipewire-0
    0 drwx------  2 chris chris 160 May 19 15:36 gnupg
    0 drwx------  3 chris chris  60 May 19 15:36 dbus-1
    0 -rw-rw----  1 chris chris   0 May 19 15:36 pipewire-0-manager.lock
    0 -rw-rw----  1 chris chris   0 May 19 15:36 pipewire-0.lock
    0 drwx------  2 chris chris  80 May 19 15:36 pulse
    0 drwx------  2 chris chris  80 May 19 15:36 keyring
    0 srwxrwxr-x  1 chris chris   0 May 19 15:36 kwallet5.socket
    0 drwx------  2 chris chris  40 May 19 15:36 gvfsd
    0 drwx------  2 chris chris  60 May 19 15:36 dconf
    0 drwx------  2 chris chris  60 May 19 15:36 at-spi
    4 -rw-rw-r--  1 chris chris  77 May 19 15:36 KSMserver__0
    4 -rw-------  1 chris chris 330 May 19 15:36 iceauth_RTpbEQ
    0 drwxr-xr-x  7 chris chris 180 May 19 15:36 systemd
    0 drwxrwxr-x  2 chris chris  60 May 19 15:36 Nextcloud
    0 drwxr-xr-x  3 root  root   60 May 19 15:36 ..
    0 drwx------  3 chris chris  80 May 19 15:39 .flatpak-helper
    4 -rw-------  1 chris chris  99 May 19 15:40 .Xauthority
    0 drwx------  3 chris chris  60 May 19 15:40 snap.thunderbird
    0 drwx------  4 chris chris  80 May 19 15:43 app
    0 drwx------  3 chris chris  60 May 19 15:51 .ecore
    0 drwxr-xr-x 10 chris chris 200 May 19 16:37 .flatpak
    0 drwxr-xr-x  2 chris chris 160 May 19 16:37 .dbus-proxy
    0 drwx------ 21 chris chris 660 May 19 16:42 .

So I think this confirms that this is an OS issue and something is 
putting /run/user/1000/doc off limits even for root though 
/run/user/1000 is open at least to root and presumably

I can see I am now way off piste here but if anyone  has any 
suggestions, even about the likely best places to take this next, I 
would appreciate it.

Many thanks Eric and all (I have learned SO much from this list over hm, 
well I have kept messages going back nine years but I know it's much 
longer back than that!)

Chris

On 19/05/2025 16:22, Eric Berger wrote:
> [forgot to hit reply-all]
> Hi Chris,
> When I have encountered similar sounding problems in the past I was 
> able to work around them as follows.
> Instead of running rstudio locally, I run rstudio-server locally and 
> then in a browser I enter the URL http://localhost:8787/
>
> HTH,
> Eric
>
>
> On Mon, May 19, 2025 at 6:21 PM Eric Berger <ericjberger using gmail.com> wrote:
>
>     Hi Chris,
>     When I have encountered similar sounding problems in the past I
>     was able to work around them as follows.
>     Instead of running rstudio locally, I run rstudio-server locally
>     and then in a browser I enter the URL http://localhost:8787/
>
>     HTH,
>     Eric
>
>     On Mon, May 19, 2025 at 6:01 PM Chris Evans via R-help
>     <r-help using r-project.org> wrote:
>
>         I realise that shiny and Rstudio is often seen as OT here but
>         I have a
>         strong suspicion that the issue is R or R/OS related so I hope
>         someone
>         here will have some thoughts that might help me.
>
>         The scenario is that as of yesterday running my shiny project
>         locally
>         from within Rtudio, i.e. by opening the index.html with
>         preview, gets me
>         "Input/output error" when I click on a link to an app.
>
>         The same shiny running on my server (https://shiny.psyctc.org)
>         runs
>         fine. So if I click on the link
>         https://shiny.psyctc.org/apps/COREpapers1 there then that app
>         runs
>         fine.  However, when I do the same locally I get "Input/output
>         error" in
>         the window that pops up where the app should run! Very
>         bizarrely, if I
>         open the app itself locally in Rstudio it runs fine.
>
>         I am completely baffled by this, have never seen it before and
>         some
>         searching on the web only suggest that "Input/output error"
>         may be an OS
>         level error and a suggestion that dmesg may give me a clue
>         about it.
>         However,dmesg is showing nothing that looks remotely pertinent
>         and
>         nothing new appears in rerunning dmesg immediately after
>         invoking that
>         error message.  I have updated and rebooted the local machine
>         (rather
>         desperate I admit) and that did update quite a lot of core
>         linux files
>         (despite my having updated this morning say six hours ago) ...
>         but
>         nothing new in R or R packages and the problem stayed the same.
>
>         To make it even more confusing the server and the local
>         machine are both
>         running R 4.5.0 (2025-04-11) and shiny version is 1.10.0 on
>         both. I use
>         git to package the server project and that gets transferred to
>         the
>         server from the local machine and commit is showing no local
>         changes and
>         the last change is showing as on 5.v.25.  The project can be
>         seen at
>         https://github.com/cpsyctc/shiny-server.
>
>         The server is running
>             Debian 5.10.197-1 (2023-09-29) x86_64 GNU/Linux
>         but the local machine runs
>             Linux tuxedo-os 6.11.0-114025-tuxedo #25~24.04.1tux1 SMP
>         PREEMPT_DYNAMIC
>
>         I would hugely appreciate it if anyone has any thoughts about
>         this, any
>         things I might do to debug it more. It's currently getting in
>         the way of
>         some mildly urgent work I need to do on the server (of course
>         it is, IT
>         knows when you are vulnerable!)
>
>         TIA,
>
>         Chris
>
>         -- 
>         Chris Evans (he/him)
>         Visiting Professor, UDLA, Quito, Ecuador
>         CORE site: http://www.coresystemtrust.org.uk
>         <https://www.coresystemtrust.org.uk/>
>         Other work web site: https://www.psyctc.org/psyctc/
>         Personal site: https://www.psyctc.org/pelerinage2016/
>         <https://ombook.psyctc.org/book>
>
>         ______________________________________________
>         R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>         https://stat.ethz.ch/mailman/listinfo/r-help
>         PLEASE do read the posting guide
>         https://www.R-project.org/posting-guide.html
>         <https://www.R-project.org/posting-guide.html>
>         and provide commented, minimal, self-contained, reproducible code.
>
-- 
Chris Evans (he/him)
Visiting Professor, UDLA, Quito, Ecuador
CORE site: http://www.coresystemtrust.org.uk 
<https://www.coresystemtrust.org.uk/>
Other work web site: https://www.psyctc.org/psyctc/
Personal site: https://www.psyctc.org/pelerinage2016/
<https://ombook.psyctc.org/book>



More information about the R-help mailing list