Failed to activate service 'org.freedesktop.timedate1' on Google Cloud VM about anytime

[eddelbuettel]

First, what version of anytime ? We do not call sysdatectl ...

edd@rob:~/git/anytime(master)$ ag timedatectl           # `ag` is a better recursive grep I use...
edd@rob:~/git/anytime(master)$ 

Second, this may be due to us trying to learn TZ and alternatives. Do you have that set? Can you play with the code in R/init.R?

There is nothing I can do for you as I do not have access to a Google Cloud VM.

[eddelbuettel]

As on my box timedatectl comes back immediately, please try on that box the simpler (command-line)

$ date                                # to see the TZ
$ Rscript -e 'format(Sys.time())'     # same

[renkun-ken]

Everything looks normal except calling timedatectl hangs.

$ date
   Tue Jul 28 11:06:16 CST 2020
$ Rscript -e 'format(Sys.time())'
[1] "2020-07-28 11:06:22"

[eddelbuettel]

Please ask Google support and your IT staff. If we learn anything, anytime can try to adjust. So far we have nothing.

You also skipped the initial questions:

• do you set TZ?
• does the system have TZ?

In the pre-systemd days I wrote an entire package trying to help guess the TZ: gettz. Do you have it installed? Would it make a difference? It is a Suggests: for anytime.

[eddelbuettel]

Also, maybe run this by hand. For me

edd@rob:~/git/anytime(master)$ timedatectl 
               Local time: Mon 2020-07-27 22:28:23 CDT 
           Universal time: Tue 2020-07-28 03:28:23 UTC 
                 RTC time: Tue 2020-07-28 03:28:23     
                Time zone: America/Chicago (CDT, -0500)
System clock synchronized: yes                         
              NTP service: active                      
          RTC in local TZ: no                          
edd@rob:~/git/anytime(master)$ 

[eddelbuettel]

R calls it, apparently, when it does not know better:

edd@rob:~/svn/r-devel$ ag timedatectl
src/library/base/man/timezones.Rd
213:  \command{timedatectl list-timezones} will list all available time zone
289:    things).  There is a command \command{timedatectl} to give details.

src/library/base/R/datetime.R
89:    ## First try timedatectl: should work on any modern Linux
91:    if (nzchar(Sys.which("timedatectl"))) {
92:        inf <- system("timedatectl", intern = TRUE)
106:                                    sQuote("timedatectl"), sQuote(tz)),

src/library/base/all.R
6083:    ## First try timedatectl: should work on any modern Linux
6085:    if (nzchar(Sys.which("timedatectl"))) {
6086:        inf <- system("timedatectl", intern = TRUE)
6100:                                    sQuote("timedatectl"), sQuote(tz)),
edd@rob:~/svn/r-devel$ 

So this goes back to my initial suggestion: 1) Do you have TZ set? 2) If not, do things change if you do?

[eddelbuettel]

@renkun-ken I definitely want to help fix this but I will need input from you as I don't have access to a system exhibiting this issue.

[renkun-ken]

@eddelbuettel Thanks for your detailed suggestions!

The TZ is set on system level before running R scripts and the hanging. However, only one VM has this issue and newly created VMs are all working well.

As the VM is rebooted, everything becomes normal. I guess it is caused by that the system service is down.

[eddelbuettel]

I guess it is caused by that the system service is down.

Phew :) Thanks for reporting back!