on_conflict returns unknown argument
new to hasura, tried looking at multiple how to on_conflict, ran mutation from api explorer and from frontend, tried upsert_users (suggest me to change it to insert)
mutation upsert_users {
insert_users(
objects: [{
auth0_id: "iexistindb",
name: "somename"}
],
on_conflict: {
constraint: users_pkey,
update_columns: [last_seen, name]
}
) {
affected_rows
}
}
expected to update the user table if auth0 already exist
so i just encountered this now. i had the on_conflict / update_columns but hadn't given update permissions to the role, only insert
Related
I am trying to figure out the difference between the admin role and the user role permissions when querying a Remote Schema, they both show that they have full access. However, when doing a query the user role cannot find one of the inputs for some reason. The query is
query SearchFacilities($getFacilitiesInput: GetFacilitiesInput!, $startDateInput: StartDateInput!) {
facilities(getFacilitiesInput: $getFacilitiesInput) {
facilityID
permitEntrances(startDateInput: $startDateInput) {
availability {
remaining
}
}
}
}
When running the query with the x-hasura-admin-secret it works fine. However, when I switch to the user role by setting a Bearer token for the user, I get the following error:
{
"errors": [
{
"extensions": {
"code": "validation-failed",
"path": "$.selectionSet.facilities.selectionSet.permitEntrances"
},
"message": "'permitEntrances' has no argument named 'startDateInput'"
}
]
}
They both have the same permissions according to the UI, in the remote schema section. Any ideas on what is causing this discrepancy? Been trying to figure this one out for a while, thanks for any help.
I am working on a messaging app using AWS AppSync.
I have the following message type...
type Message
#model
#auth(
rules: [
{ allow: groups, groups: ["externalUser"], operations: [] }
]
) {
id: ID!
channelId: ID!
senderId: ID!
channel: Channel #connection(fields: ["channelId"])
createdAt: AWSDateTime!
text: String
}
And I have a subscription onCreatemessage. I need to filter the results to only channels that the user is in. So I get a list of channels from a permissions table and add the following to my response mapping template.
$extensions.setSubscriptionFilter({
"filterGroup": [
{
"filters" : [
{
"fieldName" : "channelId",
"operator" : "in",
"value" : $context.result.channelIds
}
]
}
]
})
$util.toJson($messageResult)
And it works great. But if a user is in more than 5 channels, I get the following error.
{
"message": "Connection failed: {"errors":[{"message":"subscription exceeds maximum value limit 5 for operator `in`.","errorCode":400}]}"
}
I am new to vtl. So my question is, how can I break up that filter in to multiple or'd filters?
According to Creating enhanced subscription filters, "multiple rules in a filter are evaluated using AND logic, while multiple filters in a filter group are evaluated using OR logic".
Therefore, as I understand it, you just need to split $context.result.channelIds into groups of 5 and add an object to the filters array for each group.
Here is a VTL template that will do this for you:
#set($filters = [])
#foreach($channelId in $context.result.channelIds)
#set($group = $foreach.index / 5)
#if($filters.size() < $group + 1)
$util.qr($filters.add({
"fieldName" : "channelId",
"operator" : "in",
"value" : []
}
))
#end
$util.qr($filters.get($group).value.add($channelId))
#end
$extensions.setSubscriptionFilter({
"filterGroup": [
{
"filters" : $filters
}
]
})
You can see this template running here: https://mappingtool.dev/app/appsync/042769cd78b0e928db31212f5ee6aa17
(Note: The Mapping Tool errors on line 15 are a result of the $filters array being dynamically populated. You can safely ignore them.)
Do you want to add server-side filter for GraphQL Subscriptions?
If so, Now, Amplify is supported for server-side filter for Subscriptions.
After you checking below blog, you may sense it.
https://aws.amazon.com/blogs/mobile/announcing-server-side-filters-for-real-time-graphql-subscriptions-with-aws-amplify/
I am writing a query for Google Cloud Datastore, wanting to find the entity with a certain "moduleID" and the highest number of "sessionID_tot".
$session_query = $datastore->query()
->kind('Session')
->filter('moduleID', '=', $module)
->order('sessionID_tot', Query::ORDER_DESCENDING)
->limit(1);
$session_result = $datastore->runQuery($session_query);
I get the following error.
{ "error": { "code": 400, "message": "no matching index found.
recommended index is:\n- kind: Session\n properties:\n - name:
moduleID\n - name: sessionID_tot\n direction: desc\n", "status":
"FAILED_PRECONDITION" } }
I have read all the limitations on queries here, but can't seem to find a solution all the same. It works if I remove either the ordering or the filter for "moduleID". The properties I'm sorting and filtering are integers. Any ideas what I'm doing wrong?
As mentioned in the error, and linked to from the page you referenced. You need to create a composite index for that query.
I'm working on a React Native application with AppSync, and following is my schema to the problem:
type JoineeDeletedConnection {
items: [Joinee]
nextToken: String
}
type Mutation {
deleteJoinee(ids: [ID!]): [Joinee]
}
In 'request mapping template' to resolver to deleteJoinee, I have following (following the tutorial from https://docs.aws.amazon.com/appsync/latest/devguide/tutorial-dynamodb-batch.html):
#set($ids = [])
#foreach($id in ${ctx.args.ids})
#set($map = {})
$util.qr($map.put("id", $util.dynamodb.toString($id)))
$util.qr($ids.add($map))
#end
{
"version" : "2018-05-29",
"operation" : "BatchDeleteItem",
"tables" : {
"JoineesTable": $util.toJson($ids)
}
}
..and in 'response mapping template' to the resolver,
$util.toJson($ctx.result.data.JoineesTable)
The problem is, when I ran the query, I got empty result and nothing deleted to database as well:
// calling the query
mutation DeleteJoinee {
deleteJoinee(ids: ["xxxx", "xxxx"])
{
id
}
}
// returns
{
"data": {
"deleteJoinee": [
null
]
}
}
I finally able to solve this puzzle, thanks to the answer mentioned here to point me to some direction.
Although, I noticed that JoineesTable does have trusted entity/role to the IAM 'Roles' section, yet it wasn't working for some reason. Looking into this more, I noticed that the existing policy had following actions as default:
"Action": [
"dynamodb:DeleteItem",
"dynamodb:GetItem",
"dynamodb:PutItem",
"dynamodb:Query",
"dynamodb:Scan",
"dynamodb:UpdateItem"
]
Once I added following two more actions to the list, things have started working:
"dynamodb:BatchWriteItem",
"dynamodb:BatchGetItem"
Thanks to #Vasileios Lekakis and #Ionut Trestian on this appSync quest )
I'm having trouble displaying the created roles in a table. I'm using alanning:roles and aldeed:tabular. To create the table I have:
TabularTables.RolesAdmin = new Tabular.Table({
name: "Roles",
collection: Meteor.roles,
pub:"rolesadmin",
allow: function (userId) {
return Roles.userIsInRole(userId, 'admin');
},
columns: [
{
data: "name",
title: "Role Name",
},
],
});
And the publication looks like this:
Meteor.publish( 'rolesadmin', function() {
return Meteor.roles.find( {}, { fields: { "name": 1 } } );
});
When running the app the table only displays "Processing..." thus there is an error and it is not able at access/find the data?
I'm getting the following exception in the server terminal:
Exception from sub rolesadmin id 6c6x3mDzweP8MbB9A
Error: Did not check() all arguments during publisher 'rolesadmin'
If I check in mongo db.roles.find(), there is no role with 6c6x3mDzweP8MbB9A id. What does this error refer to?
From the meteor-tabular docs:
To tell Tabular to use your custom publish function, pass the
publication name as the pub option. Your function:
MUST accept and check three arguments: tableName, ids, and fields
MUST publish all the documents where _id is in the ids array.
MUST do any necessary security checks
SHOULD publish only the fields listed in the fields object, if one is > provided.
MAY also publish other data necessary for your table
So it looks like you'll need to account for those three arguments mentioned in the documentation appropriately. I'm not sure you actually need a custom pub for this, though, based on what you are publishing.