Sync agent configuration
An agent’s configuration can become out of sync with the configuration stored in NXLog Platform’s Agent Manager. An agent can get into this state if you update its configuration directly on the host or if the agent is offline when you deploy a configuration update from NXLog Platform. When this happens, you can instruct NXLog Platform to synchronize the agent’s configuration with the Agent Manager’s configuration or vice versa.
Prerequisites
-
A machine with NXLog Agent installed and connected to NXLog Platform.
-
Your NXLog Platform organization ID.
-
An NXLog Platform API Personal Access Token (PAT).
About the files-sync command
Calling the agents
endpoint with the files-sync
command synchronizes files between the agent and Agent Manager.
The command requires the following parameters:
-
The name of the ACL that applies to the file(s) you want to sync.
-
The file name or
*
to sync all files. -
One of the following sync actions:
-
use-agent-file-version
to send the agent’s files to Agent Manager. -
use-manager-file-version
to send the Agent Manager files to the agent. -
use-last-synced-file-version
to sync the newest files.
-
$ curl --verbose --request POST \
--url "https://agents.example.com/api/{ORG_ID}/api/v1/agents/*/files-sync/{ACL}/{FILE_NAME}/{ACTION}/?filter=({QUERY})" \
--header "Authorization: Bearer {TOKEN}" \
--data ""
If matching agents exist, the command returns 200 OK
with the following JSON body.
Otherwise, it returns an empty JSON array.
[
{
"id": "{AGENT_UUID}",
"status": "success"
}
]
Offline agents will not have their configuration updated automatically. Once they come back online, the agents will have a warning status and you must synchronize the configuration again while they are online. |
Sync configuration by agent hostname
If you need to synchronize the configuration for a single agent, the easiest way is to filter by hostname.
This example syncs the configurations of an agent with the hostname PC1
.
The hostname is case-sensitive. |
POST /agents/*/files-sync/conf/*/use-manager-file-version?filter=(hostname=PC1) | |
---|---|
Entity type |
|
Entity UUID |
|
Command |
|
ACL |
|
File |
|
Action |
|
Filter |
|
Try it
Execute the following curl command or Python script to sync an agent’s configuration by its hostname.
curl
$ curl --verbose --request POST \
--url "https://agents.example.com/api/{ORG_ID}/api/v1/agents/*/files-sync/conf/*/use-manager-file-version/?filter=(hostname=PC1)" \(1)
--header "Authorization: Bearer {TOKEN}" (2)
--data ""
1 | Replace example.com with the NXLog Platform domain you specified when installing NXLog Platform, {ORG_ID} with your organization ID, and PC1 with your agent’s hostname. |
2 | Replace {TOKEN} with your API token.
See Generating a token for instructions. |
Python
'''
Requires Python 3.x
'''
import requests
import json
# Set these variables for your environment
api_token = '<API_TOKEN>' (1)
base_url = 'https://agents.<DOMAIN>/api' (2)
org = '<ORG_ID>' (3)
agent = '<HOSTNAME>' (4)
endpoint = 'api/v1/agents'
url = '{}/{}/{}'.format(base_url, org, endpoint)
headers = {'Authorization': 'Bearer {}'.format(api_token)}
query = '*/files-sync/conf/*/use-manager-file-version?filter=(hostname={})'.format(agent)
r = requests.post('{}/{}'.format(url, query), headers=headers)
if r.status_code == 200:
print('Status: {}'.format(r.status_code))
print(json.dumps(r.json(), indent=2))
else:
print('Error: {} {}'.format(r.status_code, r.text))
1 | Replace <API_TOKEN> with your API token.
See Generating a token for instructions. |
2 | Replace <DOMAIN> with the NXLog Platform domain you specified when installing NXLog Platform. |
3 | Replace <ORG_ID> with your organization ID. |
4 | Replace <HOSTNAME> with your agent’s hostname. |
Example response
Status: 200
[
{
"id": "94fbcd8e-484c-11ef-8000-656536087e74",
"status": "success"
}
]
Update out-of-sync configurations
NXLog Platform detects when an agent’s configuration is not in sync and sets it to a warning status. You can filter for agents with a warning state and sync their configuration if the warning pertains to the configuration. This example checks for agents that are not in sync because the configuration was updated from NXLog Platform and syncs them.
GET /agents/*/?fields=id,hostname,warnings&filter=(warnings/files-modified-on-manager IS NOT NULL) | |
---|---|
Entity type |
|
Entity UUID |
|
Fields |
|
Filter |
|
Agents with an out-of-sync configuration are shown in the NXLog Platform UI with a warning status. |
POST /agents/94fbcd8e-484c-11ef-8000-656536087e74/files-sync/conf/*/use-manager-file-version | |
---|---|
Entity type |
|
Entity UUID |
|
Command |
|
ACL |
|
File |
|
Action |
|
Try it
Execute the following curl commands or Python script to update the configuration of out-of-sync agents.
curl
- List out-of-sync agents
-
$ curl --verbose --request GET \ --url "https://agents.example.com/api/{ORG_ID}/api/v1/agents/*/?fields=id,hostname,warnings&filter=(warnings%2Ffiles-modified-on-manager+IS+NOT+NULL)" \(1) --header "Authorization: Bearer {TOKEN}" (2)
1 Replace example.com
with the NXLog Platform domain you specified when installing NXLog Platform and{ORG_ID}
with your organization ID.2 Replace {TOKEN}
with your API token. See Generating a token for instructions. - Sync the configuration
-
$ curl --verbose --request POST \ --url "https://agents.example.com/api/{ORG_ID}/api/v1/agents/{AGENT_UUID}/files-sync/conf/*/use-manager-file-version" \(1) --header "Authorization: Bearer {TOKEN}" (2) --data ""
1 Replace example.com
with the NXLog Platform domain you specified when signing up,{ORG_ID}
with your organization ID, and{AGENT_UUID}
with the agent UUID you retrieved above.2 Replace {TOKEN}
with your API token. See Generating a token for instructions.
Python
'''
Requires Python 3.x
'''
import requests
import json
import re
# Set these variables for your environment
api_token = '<API_TOKEN>' (1)
base_url = 'https://agents.<DOMAIN>/api' (2)
org = '<ORG_ID>' (3)
agent = '<HOSTNAME>' (4)
endpoint = 'api/v1/agents'
url = '{}/{}/{}'.format(base_url, org, endpoint)
headers = {'Authorization': 'Bearer {}'.format(api_token)}
query = '*/?fields=id,hostname,warnings&filter=(warnings/files-modified-on-manager IS NOT NULL)'
r = requests.get('{}/{}'.format(url, query), headers=headers)
if r.status_code == 200:
acl = r'^conf.*$' (5)
for agent in r.json():
for file in agent['warnings']['filesModifiedOnManager']['files']:
if (re.match(acl, file)):
print('Syncing config for: {} - {}'.format(agent['hostname'], agent['id']))
query = '{}/files-sync/conf/*/use-manager-file-version'.format(agent['id'])
r = requests.post('{}/{}'.format(url, query), headers=headers)
if r.status_code == 200:
print('Status: {} {}'.format(r.status_code, r.reason))
else:
print('Error: {} {}'.format(r.status_code, r.text))
else:
print('Status: {} {}'.format(r.status_code, r.reason))
1 | Replace <API_TOKEN> with your API token.
See Generating a token for instructions. |
2 | Replace <DOMAIN> with the NXLog Platform domain you specified when installing NXLog Platform. |
3 | Replace <ORG_ID> with your organization ID. |
4 | Regular expression to search for files in the conf ACL. |
Example response
Retrieving out-of-sync agents with the above requests results in the following JSON array.
[
{
"id": "94fbcd8e-484c-11ef-8000-656536087e74",
"warnings": {
"filesModifiedOnManager": {
"message": "The Agent Manager changed some files, but they still need to be updated on the agent.",
"help": "Start the agent if it's offline. Wait for automatic synchronization or use the Update Configuration option to force it.",
"note": "The Agent Manager has changed some files but has not yet synchronized with the agent file contents.",
"files": [
"conf/managed.conf"
]
}
},
"hostname": "PC1"
}
]
The Python script outputs the following messages for each agent configuration it syncs.
Syncing config for: PC1 - 94fbcd8e-484c-11ef-8000-656536087e74
Status: 200 OK
Syncing config for: PC2 - 1589a98a-66b3-11ee-80d5-4f584c6f672d
Status: 200 OK