I have a bash script that I'm running from DVD. This script copies multi-volume tar files from DVD to the local machine. Part-way through the copy, the script prompts the user to insert a second DVD, at which point the remaining files are copied. The script exists on the first DVD but not on the second.
This script is simply stopping after the last file is copied, but prior to starting the tar multi-volume extract operation and subsequent processing. There are no errors or messages reported. I've tried running bash with '-x' but there's nothing suspicious - not even an exit statement. Even more unfortunate is the fact that this behavior is inconsistent. Sometimes the script will stop, but other times it will continue with no problems.
I have run strace on the script. Following the conclusion of the copy operations, I see this:
read(255, "\0\0\0\0\0\0\0\0\0\0"..., 5007) = 1302
read(255, "", 5007) = 0
exit_group(0) = ?
I know that bash reads the script file into memory and executes it from there, but is it possible that it's trying to re-read the script file at some point and failing (since it no longer exists)? The tar files are quite large, and it takes approximately 10-15 minutes from the time the script starts to the time the last file is copied (from the second DVD).
I see you have already found a workaround, so I will just try to uncover what's happening:
bash isn't reading the whole script into memory, it's doing buffered reads on it, only as much as necessary each time (presumably that's for code sharing with terminal input). Before any external commands are launched, bash seeks to the exact position in the script and continues to read from there after the command finishes. You can see this if you edit the script file while it's running:
term1$ cat > test.sh
sleep 8
echo DONE
term1$ bash test.sh
While the sleep is executing, change the script from another terminal:
term2$ cat > test.sh
echo HAHA
Observe how bash becomes confused when the sleep is complete:
test.sh: line 2: A: command not found
It remembers that the position in the input file was 8 before the sleep, so it tries to read from there and is confronted with the last A from the overwritten script.
Now to your case. Normally, having a file open from a dvd locks the drive and prohibits disk change. If you nevertheless manage to change the disk, that should definitely involve an umount which should then invalidate the script fd. That's clearly not happening according to your strace output, which is a little strange. In any case, bash won't be able to read the rest of the script.
Related
Multiple scripts are running in my Linux server which are generating huge data and I realise that it will eat all my 500GB of storage size in next 2-5 days and scripts require 10 more days to finish the process means they need more space. So most likely I am going to have a space issue problem and I will have to restart the entire process again.
Process is like this -
script1.sh content is like below
"calling an api" > /tmp/output1.txt
script2.sh content is like below
"calling an api" > /tmp/output2.txt
Executed like this -
nohup ./script1.sh & ### this create file in /tmp/output1.txt
nohup ./script2.sh & ### this create file in /tmp/output2.txt
My understand initially was, if I will follow below steps, it should work --
when scripts are running with nohup in background execute this command -
mv /tmp/output1.txt /tmp/output1.txt_bkp; touch /tmp/output1.txt
And then transfer this file /tmp/output1.txt_bkp to another server via ftp and remove it after that to get space on server and script will keep on writing in /tmp/output1.txt file.
But this assumption was wrong and script is keep on writing in /tmp/output1.txt_bkp file. I think script is writing based on inode number that is why it is keep on writing in old file.
Now the question is how to avoid space issue without killing/restart scripts?
Essentially what you're trying to do is pull a file out from under a script that's actively writing into it. I'm not sure how nohup would let you do that.
May I suggest a different approach?
Why don't you move an x number of lines from your /tmp/output[x].txt to /tmp/output[x].txt_bkp? You can do so without much trouble while your script is running and dumping stuff into /tmp/output[x].txt. That way you can free up space by shrinking your output[x] files.
Try this as a test. Open 2 terminals (or use screen) to your Linux box. Make sure both are in the same directory. Run this command in one of your terminals:
for line in `seq 1 2000000`; do echo $line >> output1.txt; done
And then run this command in the other before the first one finishes:
head -1000 output1.txt > output1.txt_bkp && sed -i '1,+999d' output1.txt
Here is what's going to happen. The first command will start producing a file that looks like this:
1
2
3
...
2000000
The second command will chop off the first 1000 lines of output1.txt and put them into output1.txt_bkp and it will do so WHILE the file is being generated.
Afterwards, look inside output1.txt and output1.txt_bkp, you will see that the former looks like this:
1001
1002
1003
1004
...
2000000
While the latter will have the first 1000 lines. You can do the same exact thing with your logs.
A word of caution: Based on your description, your box is under a heavy load from all that dumping. This may negatively impact the process outlined above.
I am currently working on project to automate a manual task in my office. We have a process that we have to re-trigger some of our ID's when they fall in repair. As part of the process, we have to extract those ID's from a oracle DB table and then put in a file on our Linux server and run the command like this-
Example file:
$cat /task/abc_YYYYMMDD_1.txt
23456
45678
...and so on
cat abc_YYYYMMDD_1.txt | scripttoprocess -args
I am using an existing java based code called 'scripttoprocess'. I can't see what's inside this code as it is encrypted( it seems) in my script. I simply go to the location where my files are present present and then use it like this:
cd /export/incoming/task
for i in `ls abc_YYYYMMDD*.txt`;do
cat $i | scripttoprocess -args
if [ $? -eq 0];then
mv $i /export/incoming/HIST/
fi
done
scripttoprocess is and existing script. I am just calling it in my own script. My script is running continuously in a loop in the background. It simply searches for abc_YYYYMMDD_1.txt file in /task directory and if it detects such a file then it starts processing the file. But I have noticed that my script starts processing the file well before it is fully written and sometime moves the file to HIST without fully processing it.
How can handle this situation. I want to be fully sure that file is completely written before I start processing it. Secondly, Is there any way to take control of the file like preparing a control file which contains list of the files which are present in the /task directory. And then I can cat this control file and pick up file names from inside of it ? Your guidance will be much appreciated.
I used
iwatch -e close_write -c "/usr/bin/pdflatex -interaction batchmode %f" document.tex
To run a command (Latex to PDF conversion) when a file (document.tex) is closed after writing to it, which you could do as well.
However, there is a caveat: This was only meant to catch manual edits to the file and failure was not critical. Therefore, this ignores the case that immediately after closing, it is opened and written again. Ask yourself if that is good enough for you.
I agree with #TenG, normally you shouldn't move a file until it is fully written. If you know for sure that the file is finished (like a file from yesterday) then you can move it safely, otherwise you can process it, but not move it. You can for example process a part of it and remember the number of processed rows so that you don't restart from scratch next time.
If you really really want to work with files that are "in progress", sometimes tail -F works for this case, but then your bash script is an ongoing process as well, not a job, and you have to manage it.
You can also check if a file is currently open (and thus unfinished) using lsof (see https://superuser.com/questions/97844/how-can-i-determine-what-process-has-a-file-open-in-linux ; check if file is open with lsof ).
Change the process, that extracts the ID's from the oracle DB table.
You can use the mv as commented by #TenG, or put something special in the file that shows the work is done:
#!/bin/bash
source file_that_runs_sqlcommands_with_credentials
output=$(your_sql_function "select * from repairjobs")
# Something more for removing them from the table and check the number of deleted records
printf "%s\nFinished\n" "${output}" >> /task/abc_YYYYMMDD_1.txt
or
#!/bin/bash
source file_that_runs_sqlcommands_with_credentials
output=$(your_sql_function "select * from repairjobs union select 'EOF' from dual")
# Something more for removing them from the table and check the number of deleted records
printf "%s\n" "${output}" >> /task/abc_YYYYMMDD_1.txt
I am trying to write a script that keeps checking if process "XXX" gets killed and if it does, show error message with text from /proc/XXX_ID/fd/1 and start again.
I am trying to make it work on custom thinclient distro where is no more packages than needed. Linux 4.6.3TS_SMP i686
I am new in scripting in Bash and I can't seem to get it working. I were googling and trying different things last two days and I moved nowhere. What am I doing wrong?
#!/bin/bash
while [ true ] ; do
process_ID="$(pgrep XXX)"
tail -f /proc/${process_ID}/fd/1 > /bin/test2.txt
Everything works till now. Now I need to somehow check if test2.txt is empty or not and if its not, use text from it as error message and start checking again. I tried something like this
if [ -s /bin/test2.txt ]
then
err="$(cat /bin/test2.txt)"
notify-send "${err}"
else
fi
done
How about this:
output_filepath=$(readlink /proc/$pid/fd/1)
while ps $pid > /dev/null
do
sleep 1
done
tail "$output_filepath"
The whole idea only works if the stdout (fd 1) of the process is redirected into a file which can be read by someone else. Such a redirect will result in /proc/<pid>/fd/1 being a symlink. We read and memorize the target of that symlink in the first line.
Then we wait until the ps of the given PID fails which typically means the process has terminated. Then we tail the file at the memorized path.
This approach has several weaknesses. The process (or its caller) could remove, modify, or rename the output file at termination time, or reading it can somehow be impossible (permissions, etc.). The output could be redirected to something not being a file (e.g. a pipe, a socket, etc.), then tailing it won't work.
ps not failing does not necessarily mean that the same process is still there. It's next to impossible that a PID is reused that quickly, but not completely impossible.
For some reasons not relevant to this question, I am running a Java server in a bash script not directly but via command substitution under a separate sub-shell, and in the background. The intent is for the subcommand to return the process id of the Java server as its standard output. The fragement in question is as follows:
launch_daemon()
{
/bin/bash <<EOF
$JAVA_HOME/bin/java $JAVA_OPTS -jar $JAR_FILE daemon $PWD/config/cl.yml <&- &
pid=\$!
echo \${pid} > $PID_FILE
echo \${pid}
EOF
}
daemon_pid=$(launch_daemon)
echo ${daemon_pid} > check.out
The Java daemon in question prints to standard error and quits if there is a problem in initialization, otherwise it closes standard out and standard err and continues on its way. Later in the script (not shown) I do a check to make sure the server process is running. Now on to the problem.
Whenever I check the $PID_FILE above, it contains the correct process id on one line.
But when I check the file check.out, it sometimes contains the correct id, other times it contains the process id repeated twice on the same line separated by a space charcater as in:
34056 34056
I am using the variable $daemon_pid in the script above later on in the script to check if the server is running, so if it contains the pid repeated twice this totally throws off the test and it incorrectly thinks the server is not running. Fiddling with the script on my server box running CentOS Linux by putting in more echo statements etc. seems to flip the behavior back to the correct one of $daemon_pid containing the process id just once, but if I think that has fixed it and check in this script to my source code repo and do a build and deploy again, I start seeing the same bad behavior.
For now I have fixed this by assuming that $daemon_pid could be bad and passing it through awk as follows:
mypid=$(echo ${daemon_pid} | awk '{ gsub(" +.*",""); print $0 }')
Then $mypid always contains the correct process id and things are fine, but needless to say I'd like to understand why it behaves the way it does. And before you ask, I have looked and looked but the Java server in question does NOT print its process id to its standard out before closing standard out.
Would really appreciate expert input.
Following the hint by #WilliamPursell, I tracked this down in the bash source code. I honestly don't know whether it is a bug or not; all I can say is that it seems like an unfortunate interaction with a questionable use case.
TL;DR: You can fix the problem by removing <&- from the script.
Closing stdin is at best questionable, not just for the reason mentioned by #JonathanLeffler ("Programs are entitled to have a standard input that's open.") but more importantly because stdin is being used by the bash process itself and closing it in the background causes a race condition.
In order to see what's going on, consider the following rather odd script, which might be called Duff's Bash Device, except that I'm not sure that even Duff would approve: (also, as presented, it's not that useful. But someone somewhere has used it in some hack. Or, if not, they will now that they see it.)
/bin/bash <<EOF
if (($1<8)); then head -n-$1 > /dev/null; fi
echo eight
echo seven
echo six
echo five
echo four
echo three
echo two
echo one
EOF
For this to work, bash and head both have to be prepared to share stdin, including sharing the file position. That means that bash needs to make sure that it flushes its read buffer (or not buffer), and head needs to make sure that it seeks back to the end of the part of the input which it uses.
(The hack only works because bash handles here-documents by copying them into a temporary file. If it used a pipe, it wouldn't be possible for head to seek backwards.)
Now, what would have happened if head had run in the background? The answer is, "just about anything is possible", because bash and head are racing to read from the same file descriptor. Running head in the background would be a really bad idea, even worse than the original hack which is at least predictable.
Now, let's go back to the actual program at hand, simplified to its essentials:
/bin/bash <<EOF
cmd <&- &
echo \$!
EOF
Line 2 of this program (cmd <&- &) forks off a separate process (to run in the background). In that process, it closes stdin and then invokes cmd.
Meanwhile, the foreground process continues reading commands from stdin (its stdin fd hasn't been closed, so that's fine), which causes it to execute the echo command.
Now here's the rub: bash knows that it needs to share stdin, so it can't just close stdin. It needs to make sure that stdin's file position is pointing to the right place, even though it may have actually read ahead a buffer's worth of input. So just before it closes stdin, it seeks backwards to the end of the current command line. [1]
If that seek happens before the foreground bash executes echo, then there is no problem. And if it happens after the foreground bash is done with the here-document, also no problem. But what if it happens while the echo is working? In that case, after the echo is done, bash will reread the echo command because stdin has been rewound, and the echo will be executed again.
And that's precisely what is happening in the OP. Sometimes, the background seek completes at just the wrong time, and causes echo \${pid} to be executed twice. In fact, it also causes echo \${pid} > $PID_FILE to execute twice, but that line is idempotent; had it been echo \${pid} >> $PID_FILE, the double execution would have been visible.
So the solution is simple: remove <&- from the server start-up line, and optionally replace it with </dev/null if you want to make sure the server can't read from stdin.
Notes:
Note 1: For those more familiar with bash source code and its expected behaviour than I am, I believe that the seek and close takes place at the end of case r_close_this: in function do_redirection_internal in redir.c, at approximately line 1093:
check_bash_input (redirector);
close_buffered_fd (redirector);
The first call does the lseek and the second one does the close. I saw the behaviour using strace -f and then searched the code for a plausible looking lseek, but I didn't go to the trouble of verifying in a debugger.
First the background to this intriguing challenge. The continuous integration build can often have failures during development and testing of deadlocks, loops, or other issues that result in a never ending test. So all the mechanisms for notifying that a build has failed become useless.
The solution will be to have the build script timeout if there's zero output to the build log file for more than 5 minutes since the build routinely writes out the names of unit tests as it proceeds. So that's the best way to identify it's "frozen".
Okay. Now the nitty gritty...
The build server uses Hudson to run a simple bash script that invokes the more complex build script based on Nant and MSBuild (all on Windows).
So far all solutions around the net involve a timeout on the total run time of the command. But that solution fails in this case because the tests might hang or freeze in the first 5 minutes.
What we've thought of so far:
First, here's the high level bash command run the full test suite in Hudson.
build.sh clean free test
That command simply sends all the Nant and MSBuild build logging to stdout.
It's obvious that we need to tee that output to a file:
build.sh clean free test 2>&1 | tee build.out
Then in parallel a command needs to sleep, check the modify time of the file and if more than 5 minutes kill the main process. A kill -9 will be fine at that point--nothing graceful needed once it has frozen.
That's the part you can help with.
In fact, I made a script like this over 15 years ago to kill the connection with a data phone line to japan after periods of inactivity but can't remember how I did it.
Sincerely,
Wayne
build.sh clean free test 2>&1 | tee build.out &
sleep 300
kill -KILL %1
You may be able to use timeout:
timeout 300 command
Solved this myself by writing a bash script.
It's called iotimeout with one parameter which is the number of seconds.
You use it like this:
build.sh clean dev test | iotimeout 120
iotimeout has 2 loops.
One is a simple while read line loop that echos echo line but
it also uses the touch command to update the modified time of a
tmp file every time it writes a line. Unfortunately, it wasn't
possible to monitor a build.out file because Windoze doesn't
update the file modified time until you close the file. Oh well.
Another loop runs in the background, that's a forever loop
which sleeps 10 seconds and then checks the modified time
of the temp file. If that ever exceeds 120 seconds old then
that loop forces the entire process group to exit.
The only tricky stuff was returning the exit code of the original
program. Bash gives you a PIPESTATUS array to solve that.
Also, figuring out how to kill the entire program group was
some research but turns out to be easy just--kill 0