fabric-docs/docs/mirroring/google-bigquery-faq.yml at 7b63df34d933565341901758a78a7a2fa613b1e8 · MicrosoftDocs/fabric-docs · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
### YamlMime:FAQ
metadata:
  title: "Frequently asked questions for Google BigQuery mirroring in Fabric"
  description: This article answers frequently asked questions about mirroring Google BigQuery in Microsoft Fabric.
  ms.reviewer: misaacs
  ms.subservice: data-warehouse
  ms.topic: faq
  ms.date: 09/09/2025
title: "Frequently asked questions for mirroring Google BigQuery in Microsoft Fabric"
summary: |
  This article answers frequently asked questions about mirroring BigQuery in Microsoft Fabric.

sections:
  - name: General
    questions:
    - question: |
       What authentication is allowed?
      answer:
        For BigQuery, we're currently supporting service key authentication. Users need to create a service key file through Google.
    - question: |
        Is SSO supported for authentication to set up replication?
      answer:
         No, we currently don't support SSO for authentication to set up replication.
    - question: |
        Why do you only support insert-only for non-primary key tables?
      answer:
        Without primary keys, there would be no way to bring in updates or deletes from your Google BigQuery Tables into your Fabric OneLake; we wouldn't know which row to update without a primary key. Insert-only for nonprimary key tables ensure data accuracy.
    - question: |
        What BigQuery user permissions are needed?
      answer:
        See our [security documentation](google-bigquery-security.md#security-considerations).
    - question: |
        What version of On-Premises Data Gateway (OPDG) is needed?
      answer:
        Version  3000.286.6 or greater is needed.
    - question: |
       Is there a staging or landing zone for BigQuery? If so, is it outside of OneLake?
      answer:
        For BigQuery, we don't use a landing zone. Instead, we utilize a direct read, which eliminates the need for staging and/or a landing zone.
    - question: |
       How long does it take to replicate inserts/updates/deletes?
      answer:
       Replication delay is roughly 15-minutes, due to limitations from Google BigQuery.
    - question: |
       Is backoff logic per table?
      answer:
       Yes, it is per table.
    - question: |
        What built-in intelligence exists for back off logic?
      answer:
       The backoff time is exponentially increased every time there's no data fetched from BigQuery or for a transient error, up to 1 hour. The frequency will automatically return to regular intervals once updated data is detected from BigQuery.
    - question: |
        How are these replicator related queries in mirroring identified?
      answer:
       Currently, we don't have a way to map those queries.
    - question: |
        What can a customer do to control backoff logic?
      answer:
       Customers don't need to do anything.
    - question: |
        Do you support replicating of views, transient or external tables?
      answer:
       No, only replicating regular tables are supported.
    - question: |
        How do I manage connections?
      answer:
       Ensure your BigQuery Connection API is connected, and you have the necessary IAM roles to view connection details. Afterwards, you can view the connection details through selecting “connection details” in the BigQuery console.
    - question: |
        What does stopping mirroring do?
      answer:
       We stop replication at the source, but we keep a copy of the tables in OneLake.

  - name: Cost management
    questions:
    - question: |
        What do we recommend a customer do to avoid or reduce BigQuery costs?
      answer:
       Implement BigQuery budgets, leverage limits on credits, or use dedicated a smaller BigQuery instance based on requirements.
    - question: |
        How are ingress fees handled?
      answer:
        Fabric doesn't charge for ingress fees into OneLake for mirroring.
    - question: |
        How are egress fees handled?
      answer:
       See BigQuery and your cloud documentation for egress fee information.

  - name: Performance
    questions:
    - question: |
        How long does the initial replication take?
      answer:
        It depends on the size of the data that is being brought in.
    - question: |
        How long does it take to replicate inserts/updates/deletes?
      answer:
        Around 15 minutes due to limitations from Google BigQuery.
    - question: |
        Will the Power BI reports use direct lake mode?
      answer:
        Yes, tables are all v-ordered delta tables.

  - name: Troubleshoot mirroring Google BigQuery in Microsoft Fabric
    questions:
      - question: |
          What are the replication statuses?
        answer:
          See [Monitor Fabric Mirror replication](monitor.md).
      - question: |
          Can mirroring for Google BigQuery be accessed through the Power BI Gateway or behind a firewall?
        answer:
          Currently, access through OPDG and VNET are supported.
      - question: |
          What does stopping/starting the mirroring do?
        answer:
          The data from source tables will be reinitialized. Each time you stop and start, the entire table is fetched again.
      - question: |
          What happens if I deselect a table from mirroring?
        answer:
          We stop mirroring that specific table and delete it from OneLake.
      - question: |
          If I delete the Mirror does it affect the source mirrored database?
        answer:
          No, we just remove the streaming tables.
      - question: |
          Can I Mirror the same database multiple times?
        answer:
          Yes, you can, but you shouldn't need to. Once the data is in Fabric, it can be shared from there.
      - question: |
          Can I Mirror specific tables from my source database?
        answer:
          Yes, specific tables can be selected during mirroring configuration.

  - name: Data governance
    questions:
      - question: |
          Is data ever leaving the customers Fabric tenant?
        answer:
          No.
      - question: |
          Is data staged outside of a customer environment?
        answer:
          No, data isn't staged outside of customer environment, it's staged in the customer's OneLake.

  - name: Licensing
    questions:
      - question: |
          What are licensing options for Fabric mirroring?
        answer:
          A Power BI Premium, Fabric Capacity, or Trial Capacity is required. For more information on licensing, see [Microsoft Fabric licenses](../enterprise/licenses.md).
      - question: |
          What are the Fabric compute costs associated with mirroring?
        answer:
          Fabric compute used to replicate your data into Fabric OneLake is free. The mirroring storage cost is free up to a limit based on capacity. For more information, see [Cost of mirroring](overview.md#cost-of-mirroring) and [Microsoft Fabric Pricing](https://azure.microsoft.com/pricing/details/microsoft-fabric/). The compute for querying data using SQL, Power BI, or Spark is charged at regular rates.

additionalContent: |

  ## Related content

  - [What is mirroring in Fabric?](overview.md)
  - [Google BigQuery connector overview](/fabric/data-factory/connector-google-bigquery-overview)