QueryTooLargeException

[HareeshaSN]

{
"detailedMessage": "Failed to interpret Gremlin query as the query is too large. Please try to reduce the number of steps and retry.",
"requestId": "dfec6edc-10ea-41cc-b55f-71ab60fc89e8",
"code": "QueryTooLargeException",
"message": "Failed to interpret Gremlin query as the query is too large. Please try to reduce the number of steps and retry."
}

[kmcginnes]

Hi @HareeshaSN.

Please include the version of Graph Explorer you are using, the Gremlin environment (Neptune version, Gremlin Server version, etc), and what operation you were trying to perform in Graph Explorer that produced that error (i.e. searching across all nodes, syncing schema, expanding node, etc).

[HareeshaSN]

Graph-Explorer version : 1.9.0
Neptune Engine version : 1.3.2.1

Getting this error when I click on "Start synchronization"

https://docs.aws.amazon.com/neptune/latest/userguide/errors-engine-codes.html

[kmcginnes]

It's possible your database has many node or edge types, which causes the query to grow.

If there is no private information in the query, could you get it so I can see what is going on?

To get the query, you will need to open the browser's developer tools for Graph Explorer.

1. In Graph Explorer go to the connections screen with the connection that fails active.
2. Open browser developer tools
3. Switch to the "Network" tab
4. Clear the existing requests (The icon that looks like 🚫)
5. In Graph Explorer start the sync process
6. In the developer tools there should be a red entry for the failed request.
7. Select the failed request
8. Switch to the "Payload" tab within the request
9. Right click the query content and select "Copy value"

[HareeshaSN]

Hi @kmcginnes,

Its sensitive data, can not share.
Getting this error for /gremlin endpoint, payload of size 8.1 MB.

[kmcginnes]

Understood. That means this might take a bit more back and forth to track down.

payload of size 8.1 MB

That seems way too large for the request size. Perhaps that is the response size? The browser developer tools don't make that super clear.

If the 8.1 MB is in fact the request size then you must have hundreds or thousands of node and/or edge labels. If that is true, then we will need to discuss how best to support databases of that scale. Graph Explorer is not tested against databases of that size.

In case you are unsure which size the 8.1 MB refers to, here is a guide.

The list of network requests has a "size" column and is the "response size" or the size of the data coming back from the server.

When a request is selected you can see the headers. There is a header value called content length that is the "request size" or the size of the query that we are sending to the server.

[HareeshaSN]

Hi @kmcginnes,

8.1 MB is request payload size of /gremlin endpoint.

Graph has 20k nodes and 180k edges.

[kmcginnes]

@HareeshaSN, Graph Explorer v1.10.1 includes a change to the Gremlin schema sync process that splits up large requests in to many smaller requests.

It's possible this resolves the issue you had. Please try the latest version and let me know.