Accessing a running process' STDOUT

I’m currently doing a lot of scripting at work. Although the primary environment is Windows, I have to manage some Linux processes as well.

One example is a series of PowerShell scripts I’ve written which abstract away various ETL tasks that we need in order to get legacy data extracted, updated and inserted into other databases.

One of the longer running processes I have is a Ruby Rake task that processes an uploaded CSV file. This can potentially take a long time, and although I could modify the script to feed progress data back, I’d like to demonstrate how to follow-up from another terminal session.

Any Rake task I write that can potentially take a while, I tend to jazz up a bit with progress details. For this current example, I am using the progress_bar gem, which writes an ASCII progress bar to the terminal along with some other useful progress information.

If I were to execute the task directly, I could watch the progress directly. Since we’re running this script remotely though, this information isn’t directly accessible.

In order to gain access to the progress info for this task, we’ll need to access the process’ STDOUT from an alternate session.

The first step is to find the PID of the process we would like:

ps aux | grep ruby

>> 1001     16544 67.7 29.4 1319420 506456 ?      Rl   13:00   3:02 ruby rake phoenix:sync_locations_from_csv

Now that we have the PID, we can pass this to strace in order to gain a bit more insight into the current process.

sudo strace -p 16544 -s 80 -e write 2>&1 | grep "write(2, \"\["

We’re redirecting the output of strace in order to further filter the results using grep. Depending on what process you’re looking to monitor you may need to adjust your grep conditions.

You can also exclude the redirect and grep entirely, but this may result in too much information to be useful ;)

This post is licensed under CC BY 4.0 by the author.

Comments powered by Disqus.