Using curl from the shell, what is the best way to discard (or detect) files that are not completely downloaded because a timeout occurred? What I'm trying to do is:
curl -m 2 --compress -o "dest/#1" "http://url/{$list}"
When a timeout occurs, the log shows it, and the part of the file that was downloaded is saved to disk:
[4/13]: http://URL/123.jpg --> dest/123.jpg
99 97984 99 97189 0 0 45469 0 0:00:02 0:00:02 --:--:-- 62500
curl: (28) Operation timed out after 2000 milliseconds with 97189 bytes received
I'm trying to either get rid of the files that were not 100% downloaded, or have them listed to attempt a resume (-C flag), later.
The best solution I have found so far is to capture the stderr of the curl call, and parse it with a combination of perl and grep to get the output file names:
curl -m 2 -o "dest/#1" "http://url/{$list}" 2>curl.out
perl -pe 's/[\t\n ]+/ /g ; s/--> /\n/g' curl.out | grep -i "Curl: (28)" | perl -pe 's/ .*//g'
Related
Hi everyone this is my first question on stackoverflow!
I'm using this software (it's a NIDS); one of its features is using socat to create a proxy that saves the traffic to a pcap.
That's the command it uses to do this: /usr/bin/socat -d OPENSSL-LISTEN:50010,cipher=HIGH,method=TLS1.2,reuseaddr,pf=ip4,fk,cert=/usr/local/owlh/src/owlhnode/conf/certs/ca.pem,verify=0 SYSTEM:"/usr/sbin/tcpdump -n -r - -s 0 -G 50 -W 100 -w /usr/local/owlh/pcaps/remote-test%d%m%Y%H%M%S.pcap not port 22"
That's what happens when using curl i try to make a request to google through the proxy:
╭─myasnik#tanuki ~/…/ossihr-poc/docker ‹master*›
╰─$ export https_proxy=https://0.0.0.0:50010/
╭─myasnik#tanuki ~/…/ossihr-poc/docker ‹master*›
╰─$ export http_proxy=https://0.0.0.0:50010/
╭─myasnik#tanuki ~/…/ossihr-poc/docker ‹master*›
╰─$ curl --proxy-insecure www.google.it
curl: (52) Empty reply from server
root#owlh-node:/# /usr/bin/socat -d OPENSSL-LISTEN:50010,cipher=HIGH,method=TLS1.2,reuseaddr,pf=ip4,fk,cert=/usr/local/owlh/src/owlhnode/conf/certs/ca.pem,verify=0 SYSTEM:"/usr/sbin/tcpdump -n -r - -s 0 -G 50 -W 100 -w /usr/local/owlh/pcaps/remote-test%d%m%Y%H%M%S.pcap not port 22"
tcpdump: unknown file format
2020/08/18 12:00:08 socat[1590] W system("/usr/sbin/tcpdump -n -r - -s 0 -G 50 -W 100 -w /usr/local/owlh/pcaps/remote-test%d%m%Y%H%M%S.pcap not port 22") returned with status 256
2020/08/18 12:00:08 socat[1590] W system(): No such file or directory
2020/08/18 12:00:08 socat[1589] E waitpid(): child 1590 exited with status 1
Thanks a lot for your help in advantage!
Here is the answer to the question, i think i misunderstood the way it was supposed to be done: https://github.com/OwlH-net/OwlH-Node/issues/47
It works ok as a single tool:
curl "someURL"
curl -o - "someURL"
but it doesn't work in a pipeline:
curl "someURL" | tr -d '\n'
curl -o - "someURL" | tr -d '\n'
it returns:
(23) Failed writing body
What is the problem with piping the cURL output? How to buffer the whole cURL output and then handle it?
This happens when a piped program (e.g. grep) closes the read pipe before the previous program is finished writing the whole page.
In curl "url" | grep -qs foo, as soon as grep has what it wants it will close the read stream from curl. cURL doesn't expect this and emits the "Failed writing body" error.
A workaround is to pipe the stream through an intermediary program that always reads the whole page before feeding it to the next program.
E.g.
curl "url" | tac | tac | grep -qs foo
tac is a simple Unix program that reads the entire input page and reverses the line order (hence we run it twice). Because it has to read the whole input to find the last line, it will not output anything to grep until cURL is finished. Grep will still close the read stream when it has what it's looking for, but it will only affect tac, which doesn't emit an error.
For completeness and future searches:
It's a matter of how cURL manages the buffer, the buffer disables the output stream with the -N option.
Example:
curl -s -N "URL" | grep -q Welcome
Another possibility, if using the -o (output file) option - the destination directory does not exist.
eg. if you have -o /tmp/download/abc.txt and /tmp/download does not exist.
Hence, ensure any required directories are created/exist beforehand, use the --create-dirs option as well as -o if necessary
The server ran out of disk space, in my case.
Check for it with df -k .
I was alerted to the lack of disk space when I tried piping through tac twice, as described in one of the other answers: https://stackoverflow.com/a/28879552/336694. It showed me the error message write error: No space left on device.
You can do this instead of using -o option:
curl [url] > [file]
So it was a problem of encoding. Iconv solves the problem
curl 'http://www.multitran.ru/c/m.exe?CL=1&s=hello&l1=1' | iconv -f windows-1251 | tr -dc '[:print:]' | ...
If you are trying something similar like source <( curl -sS $url ) and getting the (23) Failed writing body error, it is because sourcing a process substitution doesn't work in bash 3.2 (the default for macOS).
Instead, you can use this workaround.
source /dev/stdin <<<"$( curl -sS $url )"
Trying the command with sudo worked for me. For example:
sudo curl -O -k 'https url here'
note: -O (this is capital o, not zero) & -k for https url.
I had the same error but from different reason. In my case I had (tmpfs) partition with only 1GB space and I was downloading big file which finally filled all memory on that partition and I got the same error as you.
I encountered the same problem when doing:
curl -L https://packagecloud.io/golang-migrate/migrate/gpgkey | apt-key add -
The above query needs to be executed using root privileges.
Writing it in following way solved the issue for me:
curl -L https://packagecloud.io/golang-migrate/migrate/gpgkey | sudo apt-key add -
If you write sudo before curl, you will get the Failed writing body error.
For me, it was permission issue. Docker run is called with a user profile but root is the user inside the container. The solution was to make curl write to /tmp since that has write permission for all users , not just root.
I used the -o option.
-o /tmp/file_to_download
In my case, I was doing:
curl <blabla> | jq | grep <blibli>
With jq . it worked: curl <blabla> | jq . | grep <blibli>
I encountered this error message while trying to install varnish cache on ubuntu. The google search landed me here for the error (23) Failed writing body, hence posting a solution that worked for me.
The bug is encountered while running the command as root curl -L https://packagecloud.io/varnishcache/varnish5/gpgkey | apt-key add -
the solution is to run apt-key add as non root
curl -L https://packagecloud.io/varnishcache/varnish5/gpgkey | apt-key add -
The explanation here by #Kaworu is great: https://stackoverflow.com/a/28879552/198219
This happens when a piped program (e.g. grep) closes the read pipe before the previous program is finished writing the whole page. cURL doesn't expect this and emits the "Failed writing body" error.
A workaround is to pipe the stream through an intermediary program that always reads the whole page before feeding it to the next program.
I believe the more correct implementation would be to use sponge, as already suggested by #nisetama in the comments:
curl "url" | sponge | grep -qs foo
I got this error trying to use jq when I didn't have jq installed. So... make sure jq is installed if you're trying to use it.
In Bash and zsh (and perhaps other shells), you can use process substitution (Bash/zsh) to create a file on the fly, and then use that as input to the next process in the pipeline chain.
For example, I was trying to parse JSON output from cURL using jq and less, but was getting the Failed writing body error.
# Note: this does NOT work
curl https://gitlab.com/api/v4/projects/ | jq | less
When I rewrote it using process substitution, it worked!
# this works!
jq "" <(curl https://gitlab.com/api/v4/projects/) | less
Note: jq uses its 2nd argument to specify an input file
Bonus: If you're using jq like me and want to keep the colorized output in less, use the following command line instead:
jq -C "" <(curl https://gitlab.com/api/v4/projects/) | less -r
(Thanks to Kowaru for their explanation of why Failed writing body was occurring. However, their solution of using tac twice didn't work for me. I also wanted to find a solution that would scale better for large files and tries to avoid the other issues noted as comments to that answer.)
I was getting curl: (23) Failed writing body . Later I noticed that I did not had sufficient space for downloading an rpm package via curl and thats the reason I was getting issue. I freed up some space and issue for resolved.
I had the same question because of my own typo mistake:
# fails because of reasons mentioned above
curl -I -fail https://www.google.com | echo $?
curl: (23) Failed writing body
# success
curl -I -fail https://www.google.com || echo $?
I added flag -s and it did the job. eg: curl -o- -s https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.1/install.sh | bash
I have the following script hosted on Github:
https://rawgit.com/oresoftware/quicklock/master/install.sh
the contents of that file are:
#!/usr/bin/env bash
set -e;
cd "$HOME"
mkdir -p "$HOME/.quicklock/locks"
curl https://rawgit.com/oresoftware/quicklock/master/install.sh > "$HOME/.quicklock/ql.sh"
echo "To complete installation of 'quicklock' add the following line to your .bash_profile file:";
echo ". \"$HOME/.quicklock/ql.sh\"";
I download and run this script with:
curl -o- https://rawgit.com/oresoftware/quicklock/master/install.sh | bash
but I get this error:
bash: line 1: Moved: command not found
That error is killing me, I cannot figure out what is causing it. I tried curl with both the -o- option and without.
The url for raw git has changed, the error itsel is from curl.
Change rawgit.com to raw.githubusercontent.com.
Another option is to add -L to have curl follow the redirect link.
I figured this out by changing bash to bash -x. Here is the output:
curl -o- https://rawgit.com/oresoftware/quicklock/master/install.sh | bash -x
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 107 100 107 0 0 400 0 --:--:-- --:--:-- --:--:-- 402
+(:1): Moved Permanently. Redirecting to https://raw.githubusercontent.com/oresoftware/quicklock/master/install.sh
bash: line 1: Moved: command not found
#xxfelixxx is pretty much right
This was sort of nightmare, but there appears to be a redirect even when using raw.githubusercontent.com
the only thing that worked with curl was to use:
curl -o- https://raw.githubusercontent.com/oresoftware/quicklock/master/install.sh | bash
For the scripts that require arguments, you can do _ for the script placeholder and then the arguments. For exampe: example.sh that expects --help
curl -L https://raw.githubusercontent.com/<USER>/<NAME>/<BRANCH>/example.sh | bash -s _ --help
I want to download a binary from https://github.com/sschwartzman/newrelic-unix-plugin. but the url will redirect to another address, so I use the awk to parse it. e.g.
curl -I https://github.com/sschwartzman/newrelic-unix-plugin/blob/master/dist/newrelic_unix_plugin.tar.gz\?raw\=true | awk '/Location:/ {print $2}'
And the result is as I expected, e.g.
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- 0:00:01 --:--:-- 0
https://github.com/sschwartzman/newrelic-unix-plugin/raw/master/dist/newrelic_unix_plugin.tar.gz
The real address is parsed. so I just want to use curl to download it. but I always got fail.
My os is osx El Capitan (I don't have wget).
curl -I https://github.com/sschwartzman/newrelic-unix-plugin/blob/master/dist/newrelic_unix_plugin.tar.gz\?raw\=true | awk '/Location:/ {curl -O $2}'
p.s.
I tried to download the binary directly by curl, but still failed. The downloaded the file was incorrect size and content.
curl -O https://github.com/sschwartzman/newrelic-unix-plugin/raw/master/dist/newrelic_unix_plugin.tar.gz
curl has switch -L which can handle redirect.
So, this should work:
curl -L https://github.com/sschwartzman/newrelic-unix-plugin/blob/master/dist/newrelic_unix_plugin.tar.gz\?raw\=true -o newrelic_unix_plugin.tar.gz
I have a list of URLS that I need to check, to see if they still work or not. I would like to write a bash script that does that for me.
I only need the returned HTTP status code, i.e. 200, 404, 500 and so forth. Nothing more.
EDIT Note that there is an issue if the page says "404 not found" but returns a 200 OK message. It's a misconfigured web server, but you may have to consider this case.
For more on this, see Check if a URL goes to a page containing the text "404"
Curl has a specific option, --write-out, for this:
$ curl -o /dev/null --silent --head --write-out '%{http_code}\n' <url>
200
-o /dev/null throws away the usual output
--silent throws away the progress meter
--head makes a HEAD HTTP request, instead of GET
--write-out '%{http_code}\n' prints the required status code
To wrap this up in a complete Bash script:
#!/bin/bash
while read LINE; do
curl -o /dev/null --silent --head --write-out "%{http_code} $LINE\n" "$LINE"
done < url-list.txt
(Eagle-eyed readers will notice that this uses one curl process per URL, which imposes fork and TCP connection penalties. It would be faster if multiple URLs were combined in a single curl, but there isn't space to write out the monsterous repetition of options that curl requires to do this.)
wget --spider -S "http://url/to/be/checked" 2>&1 | grep "HTTP/" | awk '{print $2}'
prints only the status code for you
Extending the answer already provided by Phil. Adding parallelism to it is a no brainer in bash if you use xargs for the call.
Here the code:
xargs -n1 -P 10 curl -o /dev/null --silent --head --write-out '%{url_effective}: %{http_code}\n' < url.lst
-n1: use just one value (from the list) as argument to the curl call
-P10: Keep 10 curl processes alive at any time (i.e. 10 parallel connections)
Check the write_out parameter in the manual of curl for more data you can extract using it (times, etc).
In case it helps someone this is the call I'm currently using:
xargs -n1 -P 10 curl -o /dev/null --silent --head --write-out '%{url_effective};%{http_code};%{time_total};%{time_namelookup};%{time_connect};%{size_download};%{speed_download}\n' < url.lst | tee results.csv
It just outputs a bunch of data into a csv file that can be imported into any office tool.
This relies on widely available wget, present almost everywhere, even on Alpine Linux.
wget --server-response --spider --quiet "${url}" 2>&1 | awk 'NR==1{print $2}'
The explanations are as follow :
--quiet
Turn off Wget's output.
Source - wget man pages
--spider
[ ... ] it will not download the pages, just check that they are there. [ ... ]
Source - wget man pages
--server-response
Print the headers sent by HTTP servers and responses sent by FTP servers.
Source - wget man pages
What they don't say about --server-response is that those headers output are printed to standard error (sterr), thus the need to redirect to stdin.
The output sent to standard input, we can pipe it to awk to extract the HTTP status code. That code is :
the second ($2) non-blank group of characters: {$2}
on the very first line of the header: NR==1
And because we want to print it... {print $2}.
wget --server-response --spider --quiet "${url}" 2>&1 | awk 'NR==1{print $2}'
Use curl to fetch the HTTP-header only (not the whole file) and parse it:
$ curl -I --stderr /dev/null http://www.google.co.uk/index.html | head -1 | cut -d' ' -f2
200
wget -S -i *file* will get you the headers from each url in a file.
Filter though grep for the status code specifically.
I found a tool "webchk” written in Python. Returns a status code for a list of urls.
https://pypi.org/project/webchk/
Output looks like this:
▶ webchk -i ./dxieu.txt | grep '200'
http://salesforce-case-status.dxi.eu/login ... 200 OK (0.108)
https://support.dxi.eu/hc/en-gb ... 200 OK (0.389)
https://support.dxi.eu/hc/en-gb ... 200 OK (0.401)
Hope that helps!
Keeping in mind that curl is not always available (particularly in containers), there are issues with this solution:
wget --server-response --spider --quiet "${url}" 2>&1 | awk 'NR==1{print $2}'
which will return exit status of 0 even if the URL doesn't exist.
Alternatively, here is a reasonable container health-check for using wget:
wget -S --spider -q -t 1 "${url}" 2>&1 | grep "200 OK" > /dev/null
While it may not give you exact status out, it will at least give you a valid exit code based health responses (even with redirects on the endpoint).
Due to https://mywiki.wooledge.org/BashPitfalls#Non-atomic_writes_with_xargs_-P (output from parallel jobs in xargs risks being mixed), I would use GNU Parallel instead of xargs to parallelize:
cat url.lst |
parallel -P0 -q curl -o /dev/null --silent --head --write-out '%{url_effective}: %{http_code}\n' > outfile
In this particular case it may be safe to use xargs because the output is so short, so the problem with using xargs is rather that if someone later changes the code to do something bigger, it will no longer be safe. Or if someone reads this question and thinks he can replace curl with something else, then that may also not be safe.