Extract Instance Level Alert Expression
Hello, I am making massive number of API calls due to the high number of devices and instances in LogicMonitor to Extract Instance Level Expression for each device (VM type). I have all Global alerts but there many devices where we have Instance level alerting. Currently I am using following API to extract instance level alert expression. /device/devices/device_id/devicedatasources/devicedatasources_id/instances/instance_id/alertsettings?size=1000&filter=alertExpr!:"" I have over 26000 devices, to iterate all the devices followed by data sources and eventually instances takes millions of API calls is there any other way we can pull this information. I would need this data every month. I do not think there is something like bulk API endpoints where I can pull larger datasets in a single call. Nor there is Webhooks/Alerting Data Streams where instead of pulling it pushes updates about alert configuration at instance level. I cannot query just Device Groups as instance level alert expression are found only at Device level. Has anyone had this requirement before, any solution or alternative? Thanks Ashish
Using LogicMonitor's REST API as a Power BI Source - Part 2
Overview Back in 2020 I shared an article on how to use any of LogicMonitor's REST API methods as a datasource in Power BI Desktop for reporting. That opened a good deal of potential but also had some limitations; in particular, it relied on use of basic authentication that will eventually be deprecated, and it could only make a single API call so could only fetch up to 1,000 records. I've documented how to use a LogicMonitor bearer token instead of basic authentication, but bearer tokens aren't currently available in every portal (just our APM customers for now) and it still faces the single call limitation. In lieu of a formal Power BI data connector for LogicMonitor being available yet, there is another option available that is more secure and a good deal more flexible: using Microsoft Power BI Desktop's native support for Python! Folks familiar with LogicMonitor's APIs know there is a wealth of example Python scripts for many of our REST methods (example). These scripts not only allow us to leverage accepted methods of authentication but also allow combining calls and tweaking results in various ways that can be very useful for reporting. Best of all, using these scripts inside Power BI isn't difficult and can even be used for templated reports (I'll include some working examples at the end of this article). While these instructions focus on Power BI Desktop, reports leveraging Python can also be published to the Power BI service (Microsoft article for reference). Prerequisites Power BI Desktop. You can install this via Microsoft's Store app or from the following link:https://www.microsoft.com/en-us/download/details.aspx?id=58494 The latest version of Python installed on the same system as Power BI Desktop. This can also be installed via Microsoft's Store app or from:https://www.python.org/downloads/windows/ Some basic familiarity with LogicMonitor’s REST APIs, though I'll provide working examples here to get you started. The full API reference can be found at: https://www.logicmonitor.com/support/rest-api-developers-guide/overview/using-logicmonitors-rest-api First, install Power BI Desktop and Python, then configure each according to this Microsoft article: https://learn.microsoft.com/en-us/power-bi/connect-data/desktop-python-scripts. When adding the Python modules, you'll also want to add the 'requests' module by running: pip install requests Modifying Python Scripts for Use in Power BI As mentioned in the Microsoft articleabove, Power BI will expect Python scripts to use the Pandas module for output. If you're not familiar with Pandas, it's used extensively by the data science community for analyzing bulk data. Adding Pandas to one of the example scripts can be as easy as adding 'import pandas as pd' to the list of import statements at top of the script, then converting the JSON returned by LogicMonitor's API to a Pandas dataframe. For example, if the script captured the API results (as JSON) in a variable called "allDevices", we can convert that to a Pandas dataframe as simply as something like:pandasDevices = pd.json_normalize(allDevices) In that example, "pd" is the name we gave to the Pandas modules back in the "import" statement we added, and "json_normalize(allDevices)" tells Pandas to take our JSON and convert it to a Pandas dataframe. We can then simply print that variable as our output for Power BI Desktop to use as reporting data. Below is a full Python script that fetches all the devices from your portal and prints it as a Pandas dataframe. This is just a minor variation of an example given in our support documentation. You'd just need to enter your own LogicMonitor API ID, key, and portal name in the variables near the top. #!/bin/env python import requests import json import hashlib import base64 import time import hmac # Pandas is required for PowerBI integration... import pandas as pd # Account Info... # Your LogicMonitor portal's API access ID... AccessId = 'REPLACE_WITH_YOUR_LM_ACCESS_ID' # Your LogicMonitor portal's API access Key... AccessKey = 'REPLACE_WITH_YOUR_LM_ACCESS_KEY' # Your LogicMonitor portal. Example: if you access your portal at https://xyz.logicmonitor.com, then your portal name is "xyz"... Company = 'REPLACE_WITH_YOUR_LM_PORTAL_NAME' # Create list to keep devices... allDevices = [] # Loop through getting all devices... count = 0 done = 0 while done==0: # Request Info... httpVerb ='GET' resourcePath = '/device/devices' data='' # The following query filters for just standard on-prem resources (deviceType=0), so adjust to suite your needs... queryParams ='?v=3&offset='+str(count)+'&size=1000&fields=alertStatus,displayName,description,deviceType,id,link,hostStatus,name&filter=deviceType:0' # Construct URL... url = 'https://'+ Company +'.logicmonitor.com/santaba/rest' + resourcePath + queryParams # Get current time in milliseconds... epoch = str(int(time.time() * 1000)) # Concatenate Request details... requestVars = httpVerb + epoch + data + resourcePath # Construct signature... hmac1 = hmac.new(AccessKey.encode(),msg=requestVars.encode(),digestmod=hashlib.sha256).hexdigest() signature = base64.b64encode(hmac1.encode()) # Construct headers... auth = 'LMv1 ' + AccessId + ':' + signature.decode() + ':' + epoch headers = {'Content-Type':'application/json','Authorization':auth} # Make request... response = requests.get(url, data=data, headers=headers) # Parse response & total devices returned... parsed = json.loads(response.content) total = parsed['total'] devices = parsed['items'] allDevices.append(devices) numDevices = len(devices) count += numDevices if count == total: print ("done") done = 1 else: print ("iterating again") # (for debugging) Print all devices... # print (json.dumps(allDevices, indent=5, sort_keys=True)) # Grab just the data items... items = allDevices[0] # Convert the JSON to a Panda dataframe that PowerBI can consume... resources = pd.json_normalize(items) # Print the dataframe... print (resources) If you run that Python script directly, you'll see it prints in a columnar format instead of the raw JSON returned by the API. It's a good example of how Pandas converted the raw data into a more formal data structure for manipulation & analysis. Power BI Desktop leverages that data directly for ingestion into its own reporting engine, which is a pretty powerful combination. Now let's show how to put it to use! How to create a Power BI report that pulls data directly from LogicMonitor via Python In Power BI Desktop, click the Get Databutton. This can be to start a new report or to add to an existing report. Choose to get data from an "Other" source, choose "Python script", then click the Connect button. Paste in your complete and working Python script, then click OK. (some examples are attached to the bottom of this article) Power BI will run the script. Depending on the amount of data being retrieved this can take anywhere from a few seconds to a few minutes. When it's complete, you'll see the Navigator pane with the name of the Python Pandas dataframe from the script output. Check the box next to the item to see the data preview. If the sample looks good then click theLoadbutton. After Power BI has loaded the data you'll be presented with the report designer, ready for you to create your report. To see the full preview of the data from your portal, click theDataicon to the left of the report workspace. When you need to pull the latest data from LogicMonitor, just click theRefreshbutton in the toolbar. To Convert a Report to a Parameterized Template If you've created a Python-based report and want to save it as a re-useable, parameterized template, we'll first need to add our necessary parameters and enable Power BI to pass those values to the script. With the report active that we want to turn into a template, click theModelicon to the left of the workspace. From there, click the three dots in the upper-right corner of the table generated by the Python script and chooseEdit Query. That will open the Power Query Editor. From here clickManage Parameterson the toolbar. For our example we'll add three new parameters, which we'll call "AlertID", "AlertKey" & "PortalName" (feel free to label them however you choose). For each, enter the respective criterion used for accessing your LogicMonitor API in theCurrent Valuefield. Below's an example of what it would look like when completed. When done click theOKbutton to close the dialog. Next, click the table name in theQuerieslist ("alerts" in our example screenshot) and click theAdvanced Editoroption in the toolbar. You'll see Power BI’sM languagequery for our datasource, including the Python script embedded in it. We're going to edit the script to replace the hard-coded API parameters to use thePower BI parameters wedefined instead. Replace the values of the script'sAccessId,AccessKey, andCompanyvariables with the following, respectively (including the quotes): " & AccessID & " " & AccessKey & " " & PortalName & " Note that those will be inside the single quotes for each of the variables. Refer to the screenshot below for an example of how it would look (the changes have been highlighted). When ready click theDonebutton. ClickClose & Applyon the Power Query Editor to commit our changes. If all looks good, now let's save this as a Power BI template. Click theFilemenu, then chooseSave As. Change theSave As Typeto "Power BI template files (*.pbit)", provide a filename and clickSave. Power BI will prompt to provide a description for your template. Your report is now ready for sharing! When you open the template file, you'll be prompted to enter values for the parameters we configured in steps 2 & 3. Enter those, hit theLoadbutton, and you'll then be presented with the report designer ready to go. Example Files Here are some example Python scripts modified to use Pandas to help get you started: get_alerts.powerbi.py get_devices_powerbi.py Here are some basic, ready-to-use Power BI report templates based on the above scripts: Power BI template for reporting on Alerts Power BI template for reporting on Resources/DevicesExportable/printable reporting
LogicMonitor is great at displaying data on a screen but not so much at exporting or printing it. I’m super grateful forall the hard work my POV team put in to help me find some solutions. They really thought outside the box and we found some options to get the data out of the system that will work for us in the short term. I want to know about some of the specific goals for reporting and exporting that information from LM. I support an ecosystem that is still very reliant on having paper copies of reportingdata and having the ability to email PDF/CSV reports directly from the platform is very important to us. Even something like the ability to put a report file into Google Drive or OneDrive on a schedule would be nice. Having this be an automated process would be feature parity with the monitoring platform we are leaving.
