Hi, the excellent script that was posted here stopped working for us after upgrading to Confluence 7.19.

Apparently the GET request for the sync now needs to be a POST, with the payload being atl_token.

Replacing the last for.. loop with these lines worked for us with 7.19.6:

# extract atl_token from RESP
ATL_TOKEN=$(grep atl_token $RESP | grep value | head -1 | awk -F'"' '{print $6}')
# look for all of the available sync operations and call each one-by-one
for sync_path in $(grep /sync $RESP | grep directoryId | grep -v "{id}" | sed 's/^.*href="//; s/"//') ; do
        curl -n -k -b $COOKIE -s --header "application/x-www-form-urlencoded" -X POST --data "atl_token=${ATL_TOKEN}" "${BASEURL}${sync_path}"
done
 

Hi guys,

as a workaround please check Extender for Confluence app and documentation https://it-lab-site.atlassian.net/wiki/spaces/RAEC/pages/1904836644/REST+API+-+User+Directories 

Cheers

Adam

Hi, this a shell script version (based on the script from JRASERVER-68724) that is tested with Confluence 6.15.10.

Example Usage:

USERNAME=myusername PASSWORD=mypassword ./confluence-directory-sync.sh https://confluence.domain.com

Contents of file confluence-directory-sync.sh:

#!/bin/bash

# Source: https://jira.atlassian.com/browse/CONFSERVER-26737

#set -x

# USERNAME and PASSWORD are expected to be set as environment variables.
# The base URL of the app is given on the command line (ie: https://confluence.domain.com)
BASEURL=$1
[ -z "$BASEURL" ] && echo "usage: USERNAME=myusername PASSWORD=mypassword $0 BASEURL" && exit 1

# Required URLs and paths:
WEB_START_URL="${BASEURL}/login.action"
WEB_LOGIN_URL="${BASEURL}/dologin.action"
WEB_SUDO_URL="${BASEURL}/doauthenticate.action"
WEB_DIRLIST_PATH="/plugins/servlet/embedded-crowd/directories/list"

# store the cookie as a temp file that is deleted as we finish
COOKIE=/tmp/GetRest.$$
RESP=/tmp/GetToken.$$

# Render the login page to get some cookies
curl -k -c $COOKIE -s "${WEB_START_URL}" -o /dev/null

# login and save the cookie
curl -k -b $COOKIE -c $COOKIE -s -X POST \
 --data-urlencode "os_username=${USERNAME}" \
 --data-urlencode "os_password=${PASSWORD}" \
 --data "login='Log In'" \
 --data "os_destination=${WEB_DIRLIST_PATH}" \
 "${WEB_LOGIN_URL}" -o /dev/null

# Do web sudo
curl -k -b $COOKIE -c $COOKIE -s -X POST \
 --data-urlencode "password=${PASSWORD}" \
 --data "authenticate=Confirm" \
 --data "destination=${WEB_DIRLIST_PATH}" \
 "${WEB_SUDO_URL}" -o /dev/null

# get the list of sync opertaions by getting the URL with the cookie
curl -k -b $COOKIE -c $COOKIE -s -X GET "${BASEURL}${WEB_DIRLIST_PATH}" > $RESP

# look for all of the available sync operations and call each one-by-one
for sync_path in $(grep /sync $RESP | grep atl_token= | grep -v "{id}" | sed 's/^.*href="//; s/".$//') ; do
        curl -k -b $COOKIE -s -X GET "${BASEURL}${sync_path}"
done

# delete the temp files
rm -f $COOKIE $RESP

To trigger the directory sync for Jira Server and Bitbucket Server have a look similar scripts in the comments of these issues:

• https://jira.atlassian.com/browse/JRASERVER-68724?focusedCommentId=2394947&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-2394947
• https://jira.atlassian.com/browse/BSERV-5533?focusedCommentId=2395934&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-2395934

Hi Scott, Did you ever get to the solution with Bitbucket and JIRA?

I wrote my own because it was helpful - I'm presently digging thorough the documentation because we need it for JIRA and Bitbucket as well.

Here's the java part of the code:

{{}}

package com.keysight.confluence.support.rest;

import java.util.Map;
import java.util.List;
import javax.ws.rs.*;
import javax.ws.rs.core.MediaType;
import javax.ws.rs.core.Response;

import com.atlassian.plugins.rest.common.security.AnonymousAllowed;
import com.atlassian.crowd.manager.directory.DirectoryManager;
import com.atlassian.crowd.embedded.api.Directory;
import com.atlassian.crowd.manager.directory.SynchronisationMode;

@Path("/directory")
public class ADRestService
{
private final DirectoryManager directoryManager;

public ADRestService( DirectoryManager directoryManager)
{
this.directoryManager = directoryManager;
}

@GET
@AnonymousAllowed
@Path("sync")
public Response syncAllDirectories( ){
String response = "<p>Syncing</p>";

List directories = directoryManager.findAllDirectories();
for( Directory directory : directoryManager.findAllDirectories() ){
try{
if( directoryManager.isSynchronisable( directory.getId() )
&& !directoryManager.isSynchronising( directory.getId() )){
directoryManager.synchroniseCache( directory.getId(), SynchronisationMode.INCREMENTAL );
}
} catch( Exception e ){
// com.atlassian.crowd.exception.DirectoryInstantiationException
// com.atlassian.crowd.exception.OperationFailedException
}
}

return Response.ok( new RestResponse( response ) ).build();
}
}

It would be useful for a couple of our clients, especially during on boarding where otherwise they have to wait for a directory sync.

I can't say that a lot of customers have asked for it via support, but there have been several times where I have wished I could recommend it as an option to admins that aren't able to or don't want to decrease the interval, but want a way to easily have Confluence pull changes to group memberships (most of the time it's when they remove an employee from a permissible group and want the change effective immediately).

I don't think it will make a noticeable impact on the number support requests, however.

Adam, are a lot of customers asking for this? Would it reduce support burden?