I try to run 100 requests for some Web Service via ksh script with:
-w "total_time: %{time_total} variable. How can i count total time for all 100 requests ?
timez=0
for i in {1..100}
do
result = $(curl -s -i -H "Content-Type: application/json" -d '{"Input}": -w "total_time: %{time_total}" http//z888/zzzz/aaa/geteeeee)
done
Related
I want to create an index pattern using Opensearch API. I tried to replicate what could be made graphically in the following image window, using as index pattern name cwl-* and then as time field #timestamp.
My domain has OpenSearch 1.2 installed.
Using curl (directly modifiend the command in kibana doc):
curl -u '****:*****' -X POST "https://******.eu-central-1.es.amazonaws.com/api/index_patterns/index_pattern" -H 'osd-xsrf: true' -H 'Content-Type: application/json' -d'
{
"index_pattern": {
"title": "cwl-*",
"timeFieldName": "#timestamp"
}
}'
but I receive
{"error":{"root_cause":[{"type":"illegal_argument_exception","reason":"Rejecting mapping update to [api] as the final mapping would have more than 1 type: [_doc, index_patterns]"}],"type":"illegal_argument_exception","reason":"Rejecting mapping update to [api] as the final mapping would have more than 1 type: [_doc, index_patterns]"},"status":400}
curl -u '****:*****' -X POST "https://******.eu-central-1.es.amazonaws.com/api/index_patterns/cwl-*" -H 'osd-xsrf: true' -H 'Content-Type: application/json' -d'
{
"index_pattern": {
"title": "cwl-*",
"timeFieldName": "#timestamp"
}
}'
change api/index_patterns/index_pattern to api/index_patterns/cwl-* and try again?
It worked for me in OpenSearch 1.3 when I added an ID in the URI and used saved_objects instead of index_patterns.
So your cURL-request should work when looking like this.
curl -u '****:*****' -X POST "https://<opensearch-dashboards-host>.eu-central-1.es.amazonaws.com/api/saved_objects/index-pattern/<ID>"
-H 'osd-xsrf: true'
-H 'Content-Type: application/json'
-d
'{
"index_pattern": {
"title": "cwl-*",
"timeFieldName": "#timestamp"
}
}'
I have a curl link which is successfully run with terminal but i want to convert it as a POSTMAN request where link is,
curl -v --silent -H "Accept: application/vnd.docker.distribution.manifest.v2+json" -X GET -u <registry-user>:<registry-password> https://sc-docker-registry.eic.fullstream.ai/v2/<image-name>/manifests/<tag> 2>&1 --insecure | grep Docker-Content-Digest | awk '{print ($3)}'
I have set header and auth but unable to set 2>&1 --insecure | grep Docker-Content-Digest | awk '{print ($3)}'
Where do I set 2>&1 --insecure | grep Docker-Content-Digest | awk '{print ($3)}' command in postman
I had the issue when using "Copy all as cURL (bash)" instead of just "Copy as cURL (bash)".
A few concatenated curl commands were copied. And Postman's import didn't like that.
I had the same issue but chrome now gives you the option to copy curl as (bash) or (cmd). I had the issue when using copy as cmd but worked when I used copy as bash
I had the same error and I think it's unrelated to the pipeline arguments you sent at the end of the curl. Enclosing url part into double quotes helped with my case. Looks like its a bug in curl-to-postman library
Personally for me the problem was a semicolon at the end of the cURL request. Try to remove all the noise you can until you find what sort of character is making postman have that error
curl 'http://catalog.data.gov/api/3/' \
-H 'authority: 1fzqk3npw4.execute-api.us-east-1.amazonaws.com' \
-H 'accept: */*' \
-H 'user-agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.111 Safari/537.36' \
--compressed ;
Remove the semicolon at the end
I removed all the single quotes to Double quotes and it worked !!!
Not Working CURL
curl --location --request POST 'http://localhost/data/?type=MOBILE&value=77700023656' \
--header 'accept: */*'
WORKING CURL (changed single quotes --> double quotes)
curl --location --request POST "http://localhost/data/?type=MOBILE&value=77700023656" \
--header "accept: */*"
remove character \ in the end of all lines
for who is using vscode, replace all by regex
Can someone please help me with calculating the AWS_SIGNATURE in bash
Here is the GET I am trying to do:
curl -k \
-X GET \
-H "Host: ${AWS_BUCKET_NAME}.s3.eu-west-1.amazonaws.com" \
-H "Date: Tue, 27 Nov 2018 11:20:00 +0200" \
-H "Authorization: AWS ${AWS_ACCESS_KEY_ID}:${AWS_SIGNATURE}" \
"https://s3.eu-west-1.amazonaws.com/${AWS_BUCKET_NAME}/?list-type=2"
You need to install openssl and base64 encoder to create the signature.
Assume you supply value for each $variable
Signature=`echo -n $StringToSign | openssl sha1 -hmac $YourSecretAccessKeyID | base64`
I've a few APIs I'd like to test with cURL. I tried doing a GET as follows:
curl --user username:password --request GET http://my_domain/get_result/52d6428f3ea9a008358ad2d8/
On the server, it showed a '302' (which means redirection, right?). I'm guessing it redirected to the 'login/' page.
What is the proper way of getting this done?
Edit: I tried:
curl -c cookies.txt -b cookies.txt -L -d #login_form.txt http://my_domain/login/
where login_form.txt contains "username=username&password=password&this_is_the_login_form=1". Doesn't work. No cookies.txt files generated. And no login happening. Can you tell me how you achieve login to Django using cURL?
Here is a fully coded answer. The idea of the solution is:
you have to first visit the login page with GET to get the cookies file generated,
then parse the CSRF token out of the cookies file
and do the login using a POST request, passing the data with -d.
Afterwards you can perform any request always using that CSRF token in the data ($DJANGO_TOKEN) or with a custom X-CSRFToken header. To log out simply delete the cookies file.
Note that you need a referer (-e) to make Django's CSRF checks happy.
LOGIN_URL=https://yourdjangowebsite.com/login/
YOUR_USER='username'
YOUR_PASS='password'
COOKIES=cookies.txt
CURL_BIN="curl -s -c $COOKIES -b $COOKIES -e $LOGIN_URL"
echo -n "Django Auth: get csrftoken ..."
$CURL_BIN $LOGIN_URL > /dev/null
DJANGO_TOKEN="csrfmiddlewaretoken=$(grep csrftoken $COOKIES | sed 's/^.*csrftoken\s*//')"
echo -n " perform login ..."
$CURL_BIN \
-d "$DJANGO_TOKEN&username=$YOUR_USER&password=$YOUR_PASS" \
-X POST $LOGIN_URL
echo -n " do something while logged in ..."
$CURL_BIN \
-d "$DJANGO_TOKEN&..." \
-X POST https://yourdjangowebsite.com/whatever/
echo " logout"
rm $COOKIES
I have a slightly more secure version of this code, which uses a file for submitting the POST data, as a Gist on GitHub: django-csrftoken-login-demo.bash
Interesting background reading on Django's CSRF token is on docs.djangoproject.com.
Passing username:password in a curl request is only good for HTTP Authentication, which isn't how most websites do auth these days. Instead, you'll have to post to the login page, get the cookie, then pass it back when requesting your desired page.
Actually #Paterino answer is correct but it will not work on every implementation of sed. Instead sed 's/^.*csrftoken\s*//') we can use sed 's/^.*csrftoken[[:blank:]]*//') which is more old fashioned. MacOSXs curl doesn't use escaping, so \n\t\s don't work at all.
To use the token with a get request, use
$CURL_BIN \
-H "$DJANGO_TOKEN" \
-X GET https://yourdjangowebsite.com/whatever/
I tried using -d with -X GET, however it resulted in weird socket behaviour on the server side (Heruko H18 errors).
I'm using Django 4.1.2 and trying the #Paterino method found a couple of changes to make it work (but i have not enogh reputation to comment so wrote another answer).
Firstly, if the generated cookies.txt file is empty you have to ensure than csrf cookie is generated. I achieved this using django.views.decorators.csrf.ensure_csrf_cookie in django.contrib.auth.views.LoginView
Now, after login cookies.txt changes, so you have to recalculate DJANGO_TOKEN variable in the same way:
DJANGO_TOKEN="csrfmiddlewaretoken=$(grep csrftoken $COOKIES | sed 's/^.*csrftoken\s*//')"
From here the method doesn't change.
the accepted answer, until now(2022-12-19), has 2 issues:
misses updating DJANGO_TOKEN after login (since a new csrftoken cookie is returned after login)
doesn't include an example with a POST request (moving the csrftoken to a header) where -d already contains some payload
here is my version dealing with both:
# user and password from `./manage.py createsuperuser`
YOUR_USER='user'
YOUR_PASS='pass'
COOKIES=cookies.txt
LOGIN_URL=http://localhost:8000/admin/login/
# stores csrftoken cookie on cookies.txt
curl -s -c $COOKIES $LOGIN_URL > /dev/null
TOKEN_VALUE="$(grep -oP '(?<=csrftoken[[:space:]]).*' cookies.txt)" # https://stackoverflow.com/a/10358949/3026886 https://stackoverflow.com/a/4233691/3026886
# logs in, updating csrftoken and adding sessionid cookies
curl -b $COOKIES -c $COOKIES -d "csrfmiddlewaretoken=$TOKEN_VALUE&username=$YOUR_USER&password=$YOUR_PASS" $LOGIN_URL
# updates var env with new cookie
TOKEN_VALUE="$(grep -oP '(?<=csrftoken[[:space:]]).*' cookies.txt)"
# here comes the real request
curl -s -X POST -b $COOKIES -d "{\"a\":1}" -H "X-CSRFToken: $TOKEN_VALUE" http://localhost:8000/yourViewReceivingJsonPayload/ > /dev/null
rm cookies.txt
I've a few APIs I'd like to test with cURL. I tried doing a GET as follows:
curl --user username:password --request GET http://my_domain/get_result/52d6428f3ea9a008358ad2d8/
On the server, it showed a '302' (which means redirection, right?). I'm guessing it redirected to the 'login/' page.
What is the proper way of getting this done?
Edit: I tried:
curl -c cookies.txt -b cookies.txt -L -d #login_form.txt http://my_domain/login/
where login_form.txt contains "username=username&password=password&this_is_the_login_form=1". Doesn't work. No cookies.txt files generated. And no login happening. Can you tell me how you achieve login to Django using cURL?
Here is a fully coded answer. The idea of the solution is:
you have to first visit the login page with GET to get the cookies file generated,
then parse the CSRF token out of the cookies file
and do the login using a POST request, passing the data with -d.
Afterwards you can perform any request always using that CSRF token in the data ($DJANGO_TOKEN) or with a custom X-CSRFToken header. To log out simply delete the cookies file.
Note that you need a referer (-e) to make Django's CSRF checks happy.
LOGIN_URL=https://yourdjangowebsite.com/login/
YOUR_USER='username'
YOUR_PASS='password'
COOKIES=cookies.txt
CURL_BIN="curl -s -c $COOKIES -b $COOKIES -e $LOGIN_URL"
echo -n "Django Auth: get csrftoken ..."
$CURL_BIN $LOGIN_URL > /dev/null
DJANGO_TOKEN="csrfmiddlewaretoken=$(grep csrftoken $COOKIES | sed 's/^.*csrftoken\s*//')"
echo -n " perform login ..."
$CURL_BIN \
-d "$DJANGO_TOKEN&username=$YOUR_USER&password=$YOUR_PASS" \
-X POST $LOGIN_URL
echo -n " do something while logged in ..."
$CURL_BIN \
-d "$DJANGO_TOKEN&..." \
-X POST https://yourdjangowebsite.com/whatever/
echo " logout"
rm $COOKIES
I have a slightly more secure version of this code, which uses a file for submitting the POST data, as a Gist on GitHub: django-csrftoken-login-demo.bash
Interesting background reading on Django's CSRF token is on docs.djangoproject.com.
Passing username:password in a curl request is only good for HTTP Authentication, which isn't how most websites do auth these days. Instead, you'll have to post to the login page, get the cookie, then pass it back when requesting your desired page.
Actually #Paterino answer is correct but it will not work on every implementation of sed. Instead sed 's/^.*csrftoken\s*//') we can use sed 's/^.*csrftoken[[:blank:]]*//') which is more old fashioned. MacOSXs curl doesn't use escaping, so \n\t\s don't work at all.
To use the token with a get request, use
$CURL_BIN \
-H "$DJANGO_TOKEN" \
-X GET https://yourdjangowebsite.com/whatever/
I tried using -d with -X GET, however it resulted in weird socket behaviour on the server side (Heruko H18 errors).
I'm using Django 4.1.2 and trying the #Paterino method found a couple of changes to make it work (but i have not enogh reputation to comment so wrote another answer).
Firstly, if the generated cookies.txt file is empty you have to ensure than csrf cookie is generated. I achieved this using django.views.decorators.csrf.ensure_csrf_cookie in django.contrib.auth.views.LoginView
Now, after login cookies.txt changes, so you have to recalculate DJANGO_TOKEN variable in the same way:
DJANGO_TOKEN="csrfmiddlewaretoken=$(grep csrftoken $COOKIES | sed 's/^.*csrftoken\s*//')"
From here the method doesn't change.
the accepted answer, until now(2022-12-19), has 2 issues:
misses updating DJANGO_TOKEN after login (since a new csrftoken cookie is returned after login)
doesn't include an example with a POST request (moving the csrftoken to a header) where -d already contains some payload
here is my version dealing with both:
# user and password from `./manage.py createsuperuser`
YOUR_USER='user'
YOUR_PASS='pass'
COOKIES=cookies.txt
LOGIN_URL=http://localhost:8000/admin/login/
# stores csrftoken cookie on cookies.txt
curl -s -c $COOKIES $LOGIN_URL > /dev/null
TOKEN_VALUE="$(grep -oP '(?<=csrftoken[[:space:]]).*' cookies.txt)" # https://stackoverflow.com/a/10358949/3026886 https://stackoverflow.com/a/4233691/3026886
# logs in, updating csrftoken and adding sessionid cookies
curl -b $COOKIES -c $COOKIES -d "csrfmiddlewaretoken=$TOKEN_VALUE&username=$YOUR_USER&password=$YOUR_PASS" $LOGIN_URL
# updates var env with new cookie
TOKEN_VALUE="$(grep -oP '(?<=csrftoken[[:space:]]).*' cookies.txt)"
# here comes the real request
curl -s -X POST -b $COOKIES -d "{\"a\":1}" -H "X-CSRFToken: $TOKEN_VALUE" http://localhost:8000/yourViewReceivingJsonPayload/ > /dev/null
rm cookies.txt