Neo4j driver

[killJoy-03]

I am working on a project(.NET).

In this, we are only having read queries.
Recently we have observed our project uses maximum Memory at a certain point and gets stuck. (and memory usage is not decreasing).

So I just want to ensure we are not using neo4j driver wrong.
coreDriver is Singleton.
protected virtual IAsyncSession NewSession => coreDriver.AsyncSession(p => p.WithDefaultAccessMode(AccessMode.Read));

public async Task<IReadOnlyList<T>> ExecuteReadAsync<T>(string query, OneOf<object, IDictionary<string, object?>> parameters) where T : new() { using var session = NewSession; var res = await session.ExecuteReadAsync(async tx => { var cursor = await tx.RunAsync(query, GetOneOfValue(parameters)).ConfigureAwait(false); var res = await cursor.ToListAsync<T>().ConfigureAwait(false); return res; }).ConfigureAwait(false); return res; }

[AndyHeap-NeoTech]

Hi killJoy-03. Your usage of the session object and ExecuteRead method look correct. One thing to keep in mind is that behind the scenes when you call ToListAsync on the result cursor it is fetching results from the server as they become available and building up the complete result set in memory. If your result set is very large then it might be worth controlling the fetching of records and working with them as they arrive in a more streaming fashion.

Note: if you are not batching multiple queries into a single transaction then it is also worth considering the ExecuteQuery interface on the driver object, less code and simpler to use.

[killJoy-03]

@AndyHeap-NeoTech -- If your result set is very large then it might be worth controlling the fetching of records and working with them as they arrive in a more streaming fashion.

can you explain more like after what size and some demo code

[AndyHeap-NeoTech]

Hi KillJoy, apologies for the slow reply. I missed the notification of your message in my inbox.

With the result cursor object you can 'fetch' records as they become available, allowing you to process them as they arrive rather than building a collection in memory and taking up that space. Under the hood the ToListAsync method is doing exactly this, but fetching all of the records into memory in one go.

An example of how to do this would be:

    using var session = Driver.Session();
    return session.ExecuteRead(
            tx =>
            {
                var result = tx.Run("MATCH (a:Person) RETURN a.name ORDER BY a.name");

                //do something with each record as it is received from the server...
                result.Select(record => /*do something with the record here*/);

                //for example, build a list of the record contents as each comes in, rather than getting all records into a collection then
                //iterating through it to build the list...
                return result.Select(record => record[0].As<string>()).ToList();
            });

or you can use the FetchAsync method in a loop of your own construction.

[killJoy-03]

@AndyHeap-NeoTech Thanks for the reply and solution.