Hi,
I have been asked by a customer who asked if I can come up with a way to show the top x pages visited in the past X days.
Is there any example code out there in C# that can intercept the page info before it loads.
I can think of another way of doing it by adding some mode to each of the templates thar logs the page ID.
But I like a way I can get the Model.ID during page load.
I can work out the stat code my self. I just need a way to get the page ID.
Thanks
Darren
I know we can do it with that.
But we want a way to do it with out 3rd Party providers, like Google Analytics.
We want to built it all in to Umbraco and control all the code ourselves.
I have been thinking about it, I can add a call to the Master Template to get the ID and store that in the DB, then I just have to add some cusome code to a block grid Item and an a event that runs every 24 hours to build the stats table.
You certainly can do that, but it seems like re-inventing the wheel. Don’t forget you may need to take GDPR into account (depending on where you are) if you are storing anything that could be classed as tracking PII such as IP address.
You would need to store the request details and Umbraco page ID for every visit in a custom table (which is a database write per request), create reports as required and possibly ensure the data is not kept for longer than necessary in case the custom table grows hugely (depending on traffic).
If you’re tracking user journeys you would need a way of identifying users so storing a cookie or similar to differentiate users based on your needs.
There may be some packages that do some of what you need so possibly worth checking here first.
Just some thoughts…
Justin
Hi @darrenhunterEmerald
I completely agree with @justin-nevitech and checked the link he shared. The On My Website package is doing something which you’re trying to do. Please check it out.
Just sharing few suggestions, if you’re planning to build a custom solution. Please check the following:
For showing “Top pages” list publicly on the website or on the Backoffice, a Custom Tracker would be a nice idea, but I would recommend not save a record the exact moment a user visits a page. Under heavy traffic, this creates a bottleneck that slows down page load times.
You can use a memory buffer; so when a user visits a page, the system will instantly drop the Page ID into a temporary server memory queue (using System.Threading.Channels). This takes microseconds, allowing the page to finish loading immediately for the user.
A dedicated background worker will pull the ids from the memory queue and saves them to the database. This will completely separate the web traffic from the database writes.
You can have a Daily Cleanup task which will run every night and will group together all hits and updates a Database table which is storing the Top Pages data and then clear out the old raw data.
Be careful with the memory buffer, you can add a capacity limit to it to store some amount of hits and ignore if going beyond that.
Make sure the tracker ignores the bots and logged-in Umbraco admins otherwise stats will be inaccurate.
If you avoid storing IPs and don’t use cookies, you won’t need to add a cookie consent banner for this.
Hope this points you in the right direction!
Adding some code which might help:
using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Hosting;
using Microsoft.Extensions.Logging;
using System.Threading.Channels;
using Umbraco.Cms.Core.Composing;
using Umbraco.Cms.Core.DependencyInjection;
using Umbraco.Cms.Core.Events;
using Umbraco.Cms.Core.Notifications;
namespace MyQuickUmbracoSitetestsql10.Core
{
// 1. THE MEMORY QUEUE
public class PageTrackingQueue
{
private readonly Channel<int> _queue = Channel.CreateUnbounded<int>();
public void Enqueue(int pageId) => _queue.Writer.TryWrite(pageId);
public IAsyncEnumerable<int> DequeueAsync(CancellationToken ct) => _queue.Reader.ReadAllAsync(ct);
}
// 2. THE PAGE LOAD INTERCEPTOR
public class PageTrackingHandler : INotificationAsyncHandler<RoutingRequestNotification>
{
private readonly PageTrackingQueue _queue;
public PageTrackingHandler(PageTrackingQueue queue) => _queue = queue;
public Task HandleAsync(RoutingRequestNotification notification, CancellationToken ct)
{
var content = notification.RequestBuilder.PublishedContent;
if (content != null && content.Id > 0)
{
_queue.Enqueue(content.Id);
}
return Task.CompletedTask;
}
}
// 3. THE BACKGROUND WORKER (TEST VERSION)
public class TrackingBackgroundWorker : BackgroundService
{
private readonly PageTrackingQueue _queue;
private readonly ILogger<TrackingBackgroundWorker> _logger;
public TrackingBackgroundWorker(PageTrackingQueue queue, ILogger<TrackingBackgroundWorker> logger)
{
_queue = queue;
_logger = logger;
}
protected override async Task ExecuteAsync(CancellationToken ct)
{
await foreach (var pageId in _queue.DequeueAsync(ct))
{
// Instead of hitting a DB, we just log it to the console to prove the worker got it!
_logger.LogInformation("âś… [TRACKER TEST] Background worker successfully captured hit for Node ID: {PageId}", pageId);
}
}
}
// 4. WIRE IT UP
public class TrackingComposer : IComposer
{
public void Compose(IUmbracoBuilder builder)
{
builder.Services.AddSingleton<PageTrackingQueue>();
builder.Services.AddHostedService<TrackingBackgroundWorker>();
builder.AddNotificationAsyncHandler<RoutingRequestNotification, PageTrackingHandler>();
}
}
}
Thank you for you help, I have a look at the code in a bit
This is likely to be your biggest problem 
Hi @huwred
You can try something like this for bot check:
var userAgent = httpContext.Request.Headers.UserAgent.ToString();
if (!string.IsNullOrWhiteSpace(userAgent))
{
userAgent = userAgent.ToLowerInvariant();
string[] botKeywords =
{
"googlebot",
"bingbot",
"slurp",
"duckduckbot",
"baiduspider",
"yandexbot",
"facebookexternalhit",
"twitterbot",
"crawler",
"spider"
};
if (botKeywords.Any(bot => userAgent.Contains(bot)))
{
return Task.CompletedTask; // Drop the hit, ignore bot traffic
}
}
Honestly that’s too much 
Just wondering what’s the tech stack for this site and how’s the bot filtering logic is.
I think Cloudflare also provide something for fighting bots.
The user agent can be spoofed and set to anything, so don’t rely on checking that for bots alone. Most firewalls use AI or heuristics to detect bots.
That’s true. It needs to have multi check to solve this. having only one won’t be enough.
I don’t use cloudflare, there is a lot of filtering included some country exclusions, but if you drill down into this traffic it just looks like browser hits (mostly from the US) but it defiitely isn’t legitimate browsing 
Thanks all for your help, I built part of it that logs, Page and time and date. I need to finish building rest of the stat code.
A Bit thanks to all for help and advice.
