MongocampIndexes
The indexes of one collection with their size and how often each was used, plus creating and dropping them. It is the index card of MongocampCollectionInfos as a component of its own.
Usage
<template>
<MongocampIndexes
collection-name="orders"
:fields="['status', 'customerId', 'createdAt']"
@changed="reloadStats"
/>
</template>Props
| Prop | Type | Default | Description |
|---|---|---|---|
collectionName | string | — | Required. The component loads and reloads by itself when it changes |
canManage | boolean | true | false hides Create Index and the drop buttons (a read-only view). It is a UI hint: the server decides what is allowed |
fields | string[] | [] | Field names offered while typing a field of a new index |
Events and methods
| Name | When |
|---|---|
changed | after an index was created or dropped, so a parent can refresh its own numbers (the info cards do) |
refresh() | exposed: reloads indexes, sizes and usage |
The table
| Column | Shows |
|---|---|
| Index | the name |
| Keys | the fields in index order (country: ↑, age: ↓); a text index shows its fields as body (text) |
| Type | standard, or the badges unique, text (+ its default language), expire <seconds>, sparse, partial, hidden, collation |
| Size | bytes per index from the collection information |
| Usage | 12 uses since 11/1/2026, or a never used since … badge |
The order is the server's
The indexes come from GET …/index/details (server 1.8.1), which lists the keys as an ordered array — so the order is right for a custom-named compound index too (before 1.8.1 it only survived in the standard name c_1_b_1). Order matters: queries must filter on the first field.
A partial filter is shown as a flag only
The server lists an index's partialFilterExpression garbled ({ "n": { "underlying": null } }), so the table only says that there is a partial filter, never what it is. Creating one works.
"Never used" is a hint
Usage comes from MongoDB's $indexStats, a counter that starts again whenever the database server restarts — an index can look unused only because the server restarted recently. The _id_ index is never flagged. If the usage numbers cannot be read, the indexes are still listed with a note.
Charts
When two or more indexes have a size (or a use count), a collapsed Charts section shows the index size and the uses per index as bars, built from the numbers the table already holds (no request). Each chart has its data table as a text alternative.
Creating
Create Index opens a dialog:
| Type | Fields | Notes |
|---|---|---|
| Standard | 1–8, each ascending or descending | compound when there are several |
| Unique | 1–8 | compound unique works too; fails when existing documents repeat the values |
| Text | 1–8 | one text index may cover several fields (a collection has at most one); the direction is not used |
| Expiring | one (a date field) | Expire after: a number and s, m, h or d — 3600s, 7d |
Every kind but expiring can get a name, be sparse (only documents that have the field), hidden (kept up to date but not used by queries) and have an optional partial filter (a MongoDB filter as JSON — only matching documents are indexed; not together with sparse, the form refuses the combination). Compound order is kept: the keys go to the server as an ordered list. The form says what is missing (empty or repeated field, bad duration, a partial filter that is not a JSON object, too many fields) before anything is sent. Creating an index that already exists is a no-op on the server, not an error; an index with the same fields but other options is refused and the server's message is shown. A unique index over duplicate data answers with a readable "the existing documents repeat values of these fields" instead of MongoDB's raw error.
Dropping
The trash button asks for the index name (the same typed confirmation as dropping a collection, MongocampConfirmByName). The _id_ index has no drop button and the component refuses it as well.