Serverless Driver for Java

skinny-azure · 2024-01-08T05:22:46.009Z

Hey! Has their been any plan to implement the Neon Serverless Driver for Java? Creating a AWS Lambda Java Function and the normal JDBC connection takes 2 seconds for a mere initial connection. Looking for a way to get this down Thanks

D

dangerous-fuchsia•1/8/24, 7:36 PM

It'd be awesome to see if we could make improvements in this space, but I'm not aware of anything on our short term roadmap. I wonder if using something like Quarkus and GraalVM could improve the startup time? Or is it purely the JDBC layer that is slow?

P

precious-lavender•1/8/24, 8:24 PM

The serverless driver is just a small wrapper around the HTTP API, right? As a stop-gap solution, you could also just use the HTTP API.

Ddangerous-fuchsia It'd be awesome to see if we could make improvements in this space, but I'm not ...

S

skinny-azureOP•1/9/24, 2:30 AM

Purely JDBC

Pprecious-lavender The serverless driver is just a small wrapper around the HTTP API, right? As a s...

S

skinny-azureOP•1/9/24, 2:31 AM

I knew it existed at some point, but I couldn't find in the documentation for the life of me

S

skinny-azureOP•1/9/24, 2:33 AM

Would you mind linking the documentation to that?

P

precious-lavender•1/9/24, 2:40 AM

Let me poke the team because I can't find it either

Pprecious-lavender Let me poke the team because I can't find it either

S

skinny-azureOP•1/9/24, 2:55 AM

Would appreciate that

A

awake-maroon•1/9/24, 9:27 AM

Sorry, documentation is still in the works. Give me a few minutes and I'll give you some pointers ...

A

awake-maroon•1/9/24, 9:33 AM

The basic idea is that you https POST some JSON to the

/sql

/sql

/sql

/sql endpoint of your DB host, with a

Neon-Connection-String

Neon-Connection-String

Neon-Connection-String

Neon-Connection-String header:

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "SELECT $1 AS greeting",
    "params": [ "hello world" ]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "SELECT $1 AS greeting",
    "params": [ "hello world" ]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "SELECT $1 AS greeting",
    "params": [ "hello world" ]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "SELECT $1 AS greeting",
    "params": [ "hello world" ]
  }'

A

awake-maroon•1/9/24, 9:40 AM

This returns:

{
  "command": "SELECT",
  "fields": [
    {
      "columnID": 0,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "greeting",
      "tableID": 0
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "greeting": "hello world"
    }
  ]
}

{
  "command": "SELECT",
  "fields": [
    {
      "columnID": 0,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "greeting",
      "tableID": 0
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "greeting": "hello world"
    }
  ]
}

{
  "command": "SELECT",
  "fields": [
    {
      "columnID": 0,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "greeting",
      "tableID": 0
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "greeting": "hello world"
    }
  ]
}

{
  "command": "SELECT",
  "fields": [
    {
      "columnID": 0,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "greeting",
      "tableID": 0
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "greeting": "hello world"
    }
  ]
}

A

awake-maroon•1/9/24, 9:42 AM

There are two optional request headers to be aware of:

`Neon-Array-Mode: true` makes the `rows` object an array of arrays instead of an array of objects (e.g. the above becomes `[["hello world"]]` instead of `[{"greeting":"hello world"}]`). This saves some bandwidth, and you can still retrieve the column names from the `fields` field.
`Neon-Raw-Text-Output: true` disables parsing Postgres arrays into JSON arrays, so that the raw Postgres text format is always returned instead. For instance, try toggling `Neon-Raw-Text-Output` from `true` to `false` here:

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Raw-Text-Output: true' \
  --data '{
    "query": "SELECT $1::int[] AS pgarray",
    "params": [ [1, 2, 3] ]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Raw-Text-Output: true' \
  --data '{
    "query": "SELECT $1::int[] AS pgarray",
    "params": [ [1, 2, 3] ]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Raw-Text-Output: true' \
  --data '{
    "query": "SELECT $1::int[] AS pgarray",
    "params": [ [1, 2, 3] ]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Raw-Text-Output: true' \
  --data '{
    "query": "SELECT $1::int[] AS pgarray",
    "params": [ [1, 2, 3] ]
  }'

A

awake-maroon•1/9/24, 9:55 AM

Lastly, you can submit multiple queries in one transaction with one POST request, using the

Neon-Batch-Isolation-Level

Neon-Batch-Isolation-Level

Neon-Batch-Isolation-Level

Neon-Batch-Isolation-Level header. For instance:

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Batch-Isolation-Level: Serializable' \
  --data '{
    "queries": [
      { "query": "SELECT now() AS today", "params": [] }, 
      { "query": "SELECT now() + $1::interval AS tomorrow", "params": ["1 day"] }
    ] 
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Batch-Isolation-Level: Serializable' \
  --data '{
    "queries": [
      { "query": "SELECT now() AS today", "params": [] }, 
      { "query": "SELECT now() + $1::interval AS tomorrow", "params": ["1 day"] }
    ] 
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Batch-Isolation-Level: Serializable' \
  --data '{
    "queries": [
      { "query": "SELECT now() AS today", "params": [] }, 
      { "query": "SELECT now() + $1::interval AS tomorrow", "params": ["1 day"] }
    ] 
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  -H 'Neon-Batch-Isolation-Level: Serializable' \
  --data '{
    "queries": [
      { "query": "SELECT now() AS today", "params": [] }, 
      { "query": "SELECT now() + $1::interval AS tomorrow", "params": ["1 day"] }
    ] 
  }'

This returns:

{
  "results": [
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "today",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "today": "2024-01-09 10:00:41.670368+00"
        }
      ]
    },
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "tomorrow",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "tomorrow": "2024-01-10 10:00:41.670368+00"
        }
      ]
    }
  ]
}

{
  "results": [
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "today",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "today": "2024-01-09 10:00:41.670368+00"
        }
      ]
    },
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "tomorrow",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "tomorrow": "2024-01-10 10:00:41.670368+00"
        }
      ]
    }
  ]
}

{
  "results": [
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "today",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "today": "2024-01-09 10:00:41.670368+00"
        }
      ]
    },
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "tomorrow",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "tomorrow": "2024-01-10 10:00:41.670368+00"
        }
      ]
    }
  ]
}

{
  "results": [
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "today",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "today": "2024-01-09 10:00:41.670368+00"
        }
      ]
    },
    {
      "command": "SELECT",
      "fields": [
        {
          "columnID": 0,
          "dataTypeID": 1184,
          "dataTypeModifier": -1,
          "dataTypeSize": 8,
          "format": "text",
          "name": "tomorrow",
          "tableID": 0
        }
      ],
      "rowAsArray": false,
      "rowCount": 1,
      "rows": [
        {
          "tomorrow": "2024-01-10 10:00:41.670368+00"
        }
      ]
    }
  ]
}

A

awake-maroon•1/9/24, 9:56 AM

Obviously you can use other isolation levels, and there are also headers

Neon-Batch-Read-Only: true

Neon-Batch-Read-Only: true

Neon-Batch-Read-Only: true

Neon-Batch-Read-Only: true and

Neon-Batch-Deferrable: true

Neon-Batch-Deferrable: true

Neon-Batch-Deferrable: true

Neon-Batch-Deferrable: true available to further configure the Postgres transaction.

A

awake-maroon•1/9/24, 9:57 AM

Do note that at present the

"params"

"params"

"params"

"params" key is required, and should be an empty array if no parameters need to be passed.

A

awake-maroon•1/9/24, 10:01 AM

(Also, if you have

jq

jq

installed, pipe all my

curl

curl

examples into that for much more readable output).

Aawake-maroon The basic idea is that you https POST some JSON to the `/sql` endpoint of your D...

S

skinny-azureOP•1/10/24, 12:15 AM

Appreciate it, thanks

Aawake-maroon The basic idea is that you https POST some JSON to the `/sql` endpoint of your D...

S

skinny-azureOP•1/10/24, 12:19 AM

How does this pair in with Prepared statement efficiency?

Aawake-maroon Lastly, you can submit multiple queries in one transaction with one POST request...

S

skinny-azureOP•1/10/24, 4:29 AM

How does the response look with an

UPDATE

UPDATE

UPDATE

UPDATE ?

S

skinny-azureOP•1/11/24, 4:56 PM

@Tristan Partin

Any idea?

P

precious-lavender•1/11/24, 5:32 PM

@Sword | Andrew let me ping internally

P

precious-lavender•1/11/24, 5:34 PM

I pinged jawj internally. Note you might not hear from them until next week though

P

precious-lavender•1/11/24, 5:35 PM

I haven't interacted with the HTTP API before. I think the best you can do if you don't want to wait is read the serverless driver code, which I know is not a great answer

A

awake-maroon•1/11/24, 5:55 PM

On prepared statements: you can't use prepared statements with the HTTPS transport. I think it's rare that this would be a problem unless you're running some truly enormous joins.

A

awake-maroon•1/11/24, 6:07 PM

On an UPDATE response, I can send this:

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2",
    "params": ["X", 1]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2",
    "params": ["X", 1]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2",
    "params": ["X", 1]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2",
    "params": ["X", 1]
  }'

and get back:

{
  "command": "UPDATE",
  "fields": [],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": []
}

{
  "command": "UPDATE",
  "fields": [],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": []
}

{
  "command": "UPDATE",
  "fields": [],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": []
}

{
  "command": "UPDATE",
  "fields": [],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": []
}

Or I could use RETURNING:

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2 RETURNING *",
    "params": ["X", 1]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2 RETURNING *",
    "params": ["X", 1]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2 RETURNING *",
    "params": ["X", 1]
  }'

curl -v 'https://ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/sql' \
  -H 'Neon-Connection-String: postgres://user:password@ep-adjective-noun-nnnn.aws-region-n.aws.neon.tech/main' \
  -H 'Content-Type: application/json' \
  --data '{
    "query": "UPDATE test SET name = $1 WHERE id = $2 RETURNING *",
    "params": ["X", 1]
  }'

And get:

{
  "command": "UPDATE",
  "fields": [
    {
      "columnID": 1,
      "dataTypeID": 23,
      "dataTypeModifier": -1,
      "dataTypeSize": 4,
      "format": "text",
      "name": "id",
      "tableID": 122888
    },
    {
      "columnID": 2,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "name",
      "tableID": 122888
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "id": 1,
      "name": "X"
    }
  ]
}

{
  "command": "UPDATE",
  "fields": [
    {
      "columnID": 1,
      "dataTypeID": 23,
      "dataTypeModifier": -1,
      "dataTypeSize": 4,
      "format": "text",
      "name": "id",
      "tableID": 122888
    },
    {
      "columnID": 2,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "name",
      "tableID": 122888
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "id": 1,
      "name": "X"
    }
  ]
}

{
  "command": "UPDATE",
  "fields": [
    {
      "columnID": 1,
      "dataTypeID": 23,
      "dataTypeModifier": -1,
      "dataTypeSize": 4,
      "format": "text",
      "name": "id",
      "tableID": 122888
    },
    {
      "columnID": 2,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "name",
      "tableID": 122888
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "id": 1,
      "name": "X"
    }
  ]
}

{
  "command": "UPDATE",
  "fields": [
    {
      "columnID": 1,
      "dataTypeID": 23,
      "dataTypeModifier": -1,
      "dataTypeSize": 4,
      "format": "text",
      "name": "id",
      "tableID": 122888
    },
    {
      "columnID": 2,
      "dataTypeID": 25,
      "dataTypeModifier": -1,
      "dataTypeSize": -1,
      "format": "text",
      "name": "name",
      "tableID": 122888
    }
  ],
  "rowAsArray": false,
  "rowCount": 1,
  "rows": [
    {
      "id": 1,
      "name": "X"
    }
  ]
}

A

awake-maroon•1/11/24, 6:18 PM

@Sword | Andrew Hope that helps.

Aawake-maroon On an UPDATE response, I can send this: ```sh curl -v 'https://ep-adjective-nou...

S

skinny-azureOP•1/11/24, 6:19 PM

That exactly what I needed, appreciate it

Aawake-maroon @Sword | Andrew Hope that helps.

S

skinny-azureOP•1/12/24, 5:03 AM

Is Binary Data served in a

BinaryNode

BinaryNode

BinaryNode

BinaryNode compatible manner?

S

skinny-azureOP•1/12/24, 5:07 AM

This isn't going to change without notice, correct? Just double checking as no documentation has been released

Aawake-maroon On an UPDATE response, I can send this: ```sh curl -v 'https://ep-adjective-nou...

S

skinny-azureOP•1/12/24, 5:13 AM

Are Datatype OIDs returned?

S

skinny-azureOP•1/12/24, 7:47 AM

@jawj @Tristan Partin

{
    "command": "SELECT",
    "fields": [{
            "columnID": 2,
            "dataTypeID": 20,
            "dataTypeModifier": -1,
            "dataTypeSize": 8,
            "format": "text",
            "name": "channel_id",
            "tableID": 148344
        }, {
            "columnID": 5,
            "dataTypeID": 2950,
            "dataTypeModifier": -1,
            "dataTypeSize": 16,
            "format": "text",
            "name": "category_id",
            "tableID": 147932
        }, {
            "columnID": 6,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "display_name",
            "tableID": 147932
        }, {
            "columnID": 7,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "topic",
            "tableID": 147932
        }
    ],
    "rowAsArray": true,
    "rowCount": 1,
    "rows": [["<redacted>", null, "<redacted>", "<redacted>"]]
}

{
    "command": "SELECT",
    "fields": [{
            "columnID": 2,
            "dataTypeID": 20,
            "dataTypeModifier": -1,
            "dataTypeSize": 8,
            "format": "text",
            "name": "channel_id",
            "tableID": 148344
        }, {
            "columnID": 5,
            "dataTypeID": 2950,
            "dataTypeModifier": -1,
            "dataTypeSize": 16,
            "format": "text",
            "name": "category_id",
            "tableID": 147932
        }, {
            "columnID": 6,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "display_name",
            "tableID": 147932
        }, {
            "columnID": 7,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "topic",
            "tableID": 147932
        }
    ],
    "rowAsArray": true,
    "rowCount": 1,
    "rows": [["<redacted>", null, "<redacted>", "<redacted>"]]
}

{
    "command": "SELECT",
    "fields": [{
            "columnID": 2,
            "dataTypeID": 20,
            "dataTypeModifier": -1,
            "dataTypeSize": 8,
            "format": "text",
            "name": "channel_id",
            "tableID": 148344
        }, {
            "columnID": 5,
            "dataTypeID": 2950,
            "dataTypeModifier": -1,
            "dataTypeSize": 16,
            "format": "text",
            "name": "category_id",
            "tableID": 147932
        }, {
            "columnID": 6,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "display_name",
            "tableID": 147932
        }, {
            "columnID": 7,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "topic",
            "tableID": 147932
        }
    ],
    "rowAsArray": true,
    "rowCount": 1,
    "rows": [["<redacted>", null, "<redacted>", "<redacted>"]]
}

{
    "command": "SELECT",
    "fields": [{
            "columnID": 2,
            "dataTypeID": 20,
            "dataTypeModifier": -1,
            "dataTypeSize": 8,
            "format": "text",
            "name": "channel_id",
            "tableID": 148344
        }, {
            "columnID": 5,
            "dataTypeID": 2950,
            "dataTypeModifier": -1,
            "dataTypeSize": 16,
            "format": "text",
            "name": "category_id",
            "tableID": 147932
        }, {
            "columnID": 6,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "display_name",
            "tableID": 147932
        }, {
            "columnID": 7,
            "dataTypeID": 25,
            "dataTypeModifier": -1,
            "dataTypeSize": -1,
            "format": "text",
            "name": "topic",
            "tableID": 147932
        }
    ],
    "rowAsArray": true,
    "rowCount": 1,
    "rows": [["<redacted>", null, "<redacted>", "<redacted>"]]
}

The first field

channel_id

channel_id

channel_id

channel_id is a

bigint

bigint

in Neon. According to the JDBC

Types

Types

Types

Types Class,

BIGINT

BIGINT

BIGINT

BIGINTshould have the Id of -5; Neon returns 20.

Is their anything I'm missing?

P

precious-lavender•1/12/24, 4:29 PM

Do you know where JDBC derives these values from

P

precious-lavender•1/12/24, 4:35 PM

https://github.com/neondatabase/neon/blob/7af4c676c01bd065fff6d912420dcec542095773/proxy/src/serverless/sql_over_http.rs#L549C43-L549C48

GitHub

neon/proxy/src/serverless/sql_over_http.rs at 7af4c676c01bd065fff6d...

Neon: Serverless Postgres. We separated storage and compute to offer autoscaling, branching, and bottomless storage. - neondatabase/neon

P

precious-lavender•1/12/24, 4:35 PM

investigating!

P

precious-lavender•1/12/24, 4:51 PM

select * from pg_type where typname = 'int8';
-[ RECORD 1 ]--+---------
oid            | 20
typname        | int8
typnamespace   | 11
typowner       | 10
typlen         | 8
typbyval       | t
typtype        | b
typcategory    | N
typispreferred | f
typisdefined   | t
typdelim       | ,
typrelid       | 0
typsubscript   | -
typelem        | 0
typarray       | 1016
typinput       | int8in
typoutput      | int8out
typreceive     | int8recv
typsend        | int8send
typmodin       | -
typmodout      | -
typanalyze     | -
typalign       | d
typstorage     | p
typnotnull     | f
typbasetype    | 0
typtypmod      | -1
typndims       | 0
typcollation   | 0
typdefaultbin  | (null)
typdefault     | (null)
typacl         | (null)

select * from pg_type where typname = 'int8';
-[ RECORD 1 ]--+---------
oid            | 20
typname        | int8
typnamespace   | 11
typowner       | 10
typlen         | 8
typbyval       | t
typtype        | b
typcategory    | N
typispreferred | f
typisdefined   | t
typdelim       | ,
typrelid       | 0
typsubscript   | -
typelem        | 0
typarray       | 1016
typinput       | int8in
typoutput      | int8out
typreceive     | int8recv
typsend        | int8send
typmodin       | -
typmodout      | -
typanalyze     | -
typalign       | d
typstorage     | p
typnotnull     | f
typbasetype    | 0
typtypmod      | -1
typndims       | 0
typcollation   | 0
typdefaultbin  | (null)
typdefault     | (null)
typacl         | (null)

select * from pg_type where typname = 'int8';
-[ RECORD 1 ]--+---------
oid            | 20
typname        | int8
typnamespace   | 11
typowner       | 10
typlen         | 8
typbyval       | t
typtype        | b
typcategory    | N
typispreferred | f
typisdefined   | t
typdelim       | ,
typrelid       | 0
typsubscript   | -
typelem        | 0
typarray       | 1016
typinput       | int8in
typoutput      | int8out
typreceive     | int8recv
typsend        | int8send
typmodin       | -
typmodout      | -
typanalyze     | -
typalign       | d
typstorage     | p
typnotnull     | f
typbasetype    | 0
typtypmod      | -1
typndims       | 0
typcollation   | 0
typdefaultbin  | (null)
typdefault     | (null)
typacl         | (null)

select * from pg_type where typname = 'int8';
-[ RECORD 1 ]--+---------
oid            | 20
typname        | int8
typnamespace   | 11
typowner       | 10
typlen         | 8
typbyval       | t
typtype        | b
typcategory    | N
typispreferred | f
typisdefined   | t
typdelim       | ,
typrelid       | 0
typsubscript   | -
typelem        | 0
typarray       | 1016
typinput       | int8in
typoutput      | int8out
typreceive     | int8recv
typsend        | int8send
typmodin       | -
typmodout      | -
typanalyze     | -
typalign       | d
typstorage     | p
typnotnull     | f
typbasetype    | 0
typtypmod      | -1
typndims       | 0
typcollation   | 0
typdefaultbin  | (null)
typdefault     | (null)
typacl         | (null)

Pprecious-lavender Do you know where JDBC derives these values from

S

skinny-azureOP•1/12/24, 4:54 PM

Let me find the exact class I pulled it from

Pprecious-lavender ``` select * from pg_type where typname = 'int8'; -[ RECORD 1 ]--+--------- oid ...

S

skinny-azureOP•1/12/24, 4:54 PM

Didn't you guys give