You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In this case, we are setting the model to [“llama3.3”](https://ollama.com/library/llama3.3), a relatively new and powerful model. Remember to run `ollama run llama3.3` to pull and start the model on the server.
42
+
In this case, we are setting the model to [“llama3.2”](https://ollama.com/library/llama3.2), a relatively new and powerful model. Remember to run `ollama run llama3.2` to pull and start the model on the server.
43
43
44
44
The next json setting is the important one, overriding the endpoint that the aidb model will use.
45
45
@@ -49,12 +49,18 @@ The next json setting is the important one, overriding the endpoint that the aid
49
49
50
50
Putting those components together we get `http://llama.local:11434/v1/embeddings` as our end point.
51
51
52
-
The last JSON parameter in this example is “dimensions” which is a hint to the system about how many vector values to expect from the model. If we [look up llama3.3’s properties](https://ollama.com/library/llama3.3/blobs/4824460d29f2) we can see the `llama.embedding_length` value is 8192. The provider defaults to 1536 (with some hard-wired exceptions depending on model) but it doesn’t know about llama3.3's max. Another factor is [pgvector is limited to 2000 dimensions](https://github.com/pgvector/pgvector?tab=readme-ov-file#what-if-i-want-to-index-vectors-with-more-than-2000-dimensions). So we pass a dimension value of 2000 in the configuration, to get the maximum dimensions available with pgvector.
52
+
The last JSON parameter in this example is “dimensions” which is a hint to the system about how many vector values to expect from the model. If we [look up llama3.2’s properties](https://ollama.com/library/llama3.2/blobs/dde5aa3fc5ff) we can see the `llama.embedding_length` value is 3072. The provider defaults to 1536 (with some hard-wired exceptions depending on model) but it doesn’t know about llama3.2's embedding length. So in this case, we need to pass `"dimensions":3072` to configure aidb accordingly.
53
53
54
54
That completes the configuration parameter.
55
55
56
56
If the endpoint requires an API key, that would be passed in the credentials parameter. As this is a local model, we don’t need to pass any credentials.
57
57
58
+
59
+
!!!Note Note
60
+
When using indexing in pgvector, consider the [pgvector indexing limitations](https://github.com/pgvector/pgvector?tab=readme-ov-file#what-if-i-want-to-index-vectors-with-more-than-2000-dimensions). Aidb does not automatically configure an index today but if you add one manually, make sure it supports the number of dimensions your model uses.
61
+
62
+
!!!
63
+
58
64
## Using the model
59
65
60
66
Use the model name you created earlier to use the model just like any other Pipelines model. Here is an example of how to use the model to get an embedding:
Copy file name to clipboardExpand all lines: advocacy_docs/edb-postgres-ai/cloud-service/references/distributed_high_availability/pgd_cli_ba.mdx
+34-40
Original file line number
Diff line number
Diff line change
@@ -6,21 +6,22 @@ redirects:
6
6
- /biganimal/latest/using_cluster/pgd_cli_ba/ #generated for BigAnimal URL path removal branch
7
7
---
8
8
9
-
When running a distributed high-availability cluster on Cloud Service, you can use the [PGD CLI](/pgd/latest/cli/) to manage cluster operations. Examples of these operations include switching over write leaders, performing cluster health checks, and viewing various details about nodes, groups, or other aspects of the cluster.
9
+
When running a distributed high-availability cluster on Cloud Service, you can use the [PGD CLI](/pgd/latest/cli/) to manage cluster operations.
10
+
Examples of these operations include switching over write leaders, performing cluster health checks, and viewing various details about nodes, groups, or other aspects of the cluster.
10
11
11
12
## Installing the PGD CLI
12
13
13
14
To [install the PGD CLI](/pgd/latest/cli/installing/), for Debian and Ubuntu machines, replace `<your-token>` with your EDB subscription token in the following command:
@@ -48,70 +49,63 @@ Without a properly configured `.pgpass`, you receive a database connection error
48
49
To use the PGD CLI with your database connection string, use the `--dsn` flag with your PGD CLI command:
49
50
50
51
```bash
51
-
pgd show-nodes --dsn "<your_connection_string>"
52
+
pgd nodes list --dsn "<your_connection_string>"
52
53
```
53
54
54
55
## PGD commands in Cloud Service
55
56
56
-
!!!Note
57
-
58
-
Three EDB Postgres Distributed CLI commands don't work with distributed high-availability Cloud Service clusters: `create-proxy`, `delete-proxy`, and `alter-proxy-option`. These commands are managed by Cloud Service, as Cloud Service runs on Kubernetes. It's a technical best practice to have the Kubernetes operator handle these functions.
59
-
!!!
60
-
61
57
The examples that follow show the most common PGD CLI commands with a Cloud Service cluster.
62
58
63
-
### `pgd check-health`
59
+
### `pgd cluster show --health`
64
60
65
-
`pgd check-health` provides statuses with relevant messaging regarding the clock skew of node pairs, node accessibility, the current raft leader, replication slot health, and versioning consistency:
61
+
`pgd cluster show --health` provides statuses with relevant messaging regarding the clock skew of node pairs, node accessibility, the current raft leader, replication slot health, and versioning consistency:
Replication Slots Ok All PGD replication slots are working correctly
71
+
Clock Skew Ok Clock drift is within permissible limit
72
+
Versions Ok All nodes are running the same PGD version```
78
73
79
-
### `pgd show-nodes`
74
+
### `pgd nodes list`
80
75
81
-
`pgd show-nodes` returns all the nodes in the distributed high-availability cluster and their summaries, including name, node id, group, and current/target state:
76
+
`pgd nodes list` returns all the nodes in the distributed high-availability cluster and their summaries, including name, node id, group, and current/target state:
`pgd show-groups` returns all groups in your distributed high-availability Cloud Service cluster. It also notes the node that's the current write leader of each group:
90
+
`pgd groups show` returns all groups in your distributed high-availability Cloud Service cluster.
world 3239291720 global true true p-mbx2p83u9n-a-1
103
-
dc1 4269540889 data p-mbx2p83u9n-a false false
104
-
p-mbx2p83u9n-a 2800873689 data world true true p-mbx2p83u9n-a-3
95
+
Group Name Parent Group Name Group Type Nodes
96
+
-------------- ----------------- ---------- -----
97
+
world global 0
98
+
p-w75f4ib1pu-a world data 3
105
99
```
106
100
107
-
### `pgd switchover`
101
+
### `pgd group set-leader`
108
102
109
-
`pgd switchover` manually changes the write leader of the group and can be used to simulate a [failover](/pgd/latest/quickstart/further_explore_failover).
103
+
`pgd group set-leader` manually changes the write leader of the group and can be used to simulate a [failover](/pgd/latest/quickstart/further_explore_failover).
110
104
111
105
```
112
-
$ pgd switchover --group-name world --node-name p-mbx2p83u9n-a-2 --dsn "postgres://edb_admin@p-mbx2p83u9n-a.pg.biganimal.io:5432/bdrdb?sslmode=require"
106
+
pgd group p-w75f4ib1pu-a set-leader p-w75f4ib1pu-a-2 --dsn "postgres://edb_admin@p-w75f4ib1pu-a.vmk31wilqpjeopka.biganimal.io:5432/bdrdb?sslmode=require"
113
107
__OUTPUT__
114
-
switchover is complete
108
+
Command executed successfully
115
109
```
116
110
117
111
See the [PGD CLI command reference](/pgd/latest/cli/command_ref/) for the full range of PGD CLI commands and their descriptions.
Copy file name to clipboardExpand all lines: advocacy_docs/edb-postgres-ai/cloud-service/references/supported_cluster_types/distributed_highavailability.mdx
+1-1
Original file line number
Diff line number
Diff line change
@@ -19,7 +19,7 @@ The witness node/witness group doesn't host data but exists for management purpo
19
19
20
20
!!!Note
21
21
22
-
Operations against a distributed high-availability cluster leverage the [EDB Postgres Distributed switchover](/pgd/latest/cli/command_ref/pgd_switchover/) feature, which provides subsecond interruptions during planned lifecycle operations.
22
+
Operations against a distributed high-availability cluster leverage the [EDB Postgres Distributed set-leader](/pgd/latest/cli/command_ref/group/set-leader) feature, which provides subsecond interruptions during planned lifecycle operations.
Copy file name to clipboardExpand all lines: product_docs/docs/epas/17/database_administration/01_configuration_parameters/03_configuration_parameters_by_functionality/02_resource_usage_memory.mdx
+1-1
Original file line number
Diff line number
Diff line change
@@ -16,7 +16,7 @@ These configuration parameters control resource use pertaining to memory.
0 commit comments