Skip to content

Commit

Permalink
Release new docs
Browse files Browse the repository at this point in the history
  • Loading branch information
Milvus-doc-bot authored and Milvus-doc-bot committed Sep 19, 2024
1 parent 26d0fd8 commit 9511f84
Showing 1 changed file with 3 additions and 3 deletions.
6 changes: 3 additions & 3 deletions v2.4.x/site/en/userGuide/clustering-compaction.md
Original file line number Diff line number Diff line change
Expand Up @@ -61,8 +61,8 @@ common:
| `enable` | Specifies whether to enable clustering compaction.<br>Setting this to `true` if you need to enable this feature for every collection having a clustering key. | `false` |
| `autoEnable` | Specifies whether to enable automatically triggered compaction.<br>Setting this to `true` indicates that Milvus compacts the collections having a clustering key at the specified intervals. | `false` |
| `triggerInterval` | Specifies the interval in milliseconds at which Milvus starts clustering compaction.<br>This parameter is valid only when `autoEnable` is set to `true`. | - |
| `minInterval` | Specifies the minimum interval in milliseconds.<br>This parameter is valid only when `autoEnable` is set to `true`.<br>Setting this to an integer greater than triggerInterval helps avoid repeated compactions within a short period. | - |
| `maxInterval` | Specifies the maximum interval in milliseconds.<br>This parameter is valid only when `autoEnable` is set to `true`.<br>Once Milvus detects that a collection has not been clustering-compacted for a duration longer than this value, it forces a clustering compaction. | - |
| `minInterval` | Specifies the minimum interval in seconds.<br>This parameter is valid only when `autoEnable` is set to `true`.<br>Setting this to an integer greater than triggerInterval helps avoid repeated compactions within a short period. | - |
| `maxInterval` | Specifies the maximum interval in seconds.<br>This parameter is valid only when `autoEnable` is set to `true`.<br>Once Milvus detects that a collection has not been clustering-compacted for a duration longer than this value, it forces a clustering compaction. | - |
| `newDataSizeThreshold` | Specifies the upper threshold to trigger a clustering compaction.<br>This parameter is valid only when `autoEnable` is set to `true`.<br>Once Milvus detects that the data volume in a collection exceeds this value, it initiates a clustering compaction process. | - |
| `timeout` | Specifies the timeout duration for a clustering compaction.<br>A clustering compaction fails if its execution time exceeds this value. | - |

Expand Down Expand Up @@ -214,4 +214,4 @@ Here are some tips for you to use clustering compaction efficiently:
- Use the partition key as the clustering key.
You can set `common.usePartitionKeyAsClusteringKey` to true if you want to enable this feature for all collections in your Milvus instance or if you still face performance issues in a large collection with a partition key. By doing so, you will have a clustering key and a partition key when you choose a scalar field in a collection as the partition key.

Note that this setting does not prevent you from choosing another scalar field as the clustering key. The explicitly designated clustering key always takes precedence.
Note that this setting does not prevent you from choosing another scalar field as the clustering key. The explicitly designated clustering key always takes precedence.

0 comments on commit 9511f84

Please sign in to comment.