Updating Parts of Documents | Apache Solr Reference Guide 8.11

Once you have indexed the content you need in your Solr index, you will want to start thinking about your strategy for dealing with changes to those documents. Solr supports three approaches to updating documents that have only partially changed.

The first is atomic updates. This approach allows changing only one or more fields of a document without having to reindex the entire document.

The second approach is known as in-place updates. This approach is similar to atomic updates (is a subset of atomic updates in some sense), but can be used only for updating single valued non-indexed and non-stored docValue-based numeric fields.

The third approach is known as optimistic concurrency or optimistic locking. It is a feature of many NoSQL databases, and allows conditional updating a document based on its version. This approach includes semantics and rules for how to deal with version matches or mis-matches.

Atomic Updates (and in-place updates) and Optimistic Concurrency may be used as independent strategies for managing changes to documents, or they may be combined: you can use optimistic concurrency to conditionally apply an atomic update.

Atomic Updates

Solr supports several modifiers that atomically update values of a document. This allows updating only specific fields, which can help speed indexing processes in an environment where speed of index additions is critical to the application.

To use atomic updates, add a modifier to the field that needs to be updated. The content can be updated, added to, or incrementally increased if the field has a numeric type.

set

Set or replace the field value(s) with the specified value(s), or remove the values if 'null' or empty list is specified as the new value.

May be specified as a single value, or as a list for multiValued fields.

add

Adds the specified values to a multiValued field. May be specified as a single value, or as a list.

add-distinct

Adds the specified values to a multiValued field, only if not already present. May be specified as a single value, or as a list.

remove

Removes (all occurrences of) the specified values from a multiValued field. May be specified as a single value, or as a list.

removeregex

Removes all occurrences of the specified regex from a multiValued field. May be specified as a single value, or as a list.

inc

Increments a numeric value by a specific amount. Must be specified as a single numeric value.

Field Storage

The core functionality of atomically updating a document requires that all fields in your schema must be configured as stored (stored="true") or docValues (docValues="true") except for fields which are <copyField/> destinations, which must be configured as stored="false". Atomic updates are applied to the document represented by the existing stored field values. All data in copyField destinations fields must originate from ONLY copyField sources.

If <copyField/> destinations are configured as stored, then Solr will attempt to index both the current value of the field as well as an additional copy from any source fields. If such fields contain some information that comes from the indexing program and some information that comes from copyField, then the information which originally came from the indexing program will be lost when an atomic update is made.

There are other kinds of derived fields that must also be set so they aren’t stored. Some spatial field types, such as BBoxField and LatLonType, use derived fields. CurrencyFieldType also uses derived fields. These types create additional fields which are normally specified by a dynamic field definition. That dynamic field definition must be not stored, or indexing will fail.

Example Updating Part of a Document

If the following document exists in our collection:

{"id":"mydoc",
 "price":10,
 "popularity":42,
 "categories":["kids"],
 "sub_categories":["under_5","under_10"],
 "promo_ids":["a123x"],
 "tags":["free_to_try","buy_now","clearance","on_sale"]
}

And we apply the following update command:

{"id":"mydoc",
 "price":{"set":99},
 "popularity":{"inc":20},
 "categories":{"add":["toys","games"]},
 "sub_categories":{"add-distinct":"under_10"},
 "promo_ids":{"remove":"a123x"},
 "tags":{"remove":["free_to_try","on_sale"]}
}

The resulting document in our collection will be:

{"id":"mydoc",
 "price":99,
 "popularity":62,
 "categories":["kids","toys","games"],
 "sub_categories":["under_5","under_10"],
 "tags":["buy_now","clearance"]
}

Updating Child Documents

Solr supports modifying, adding and removing child documents as part of atomic updates. Syntactically, updates changing the children of a document are very similar to regular atomic updates of simple fields, as demonstrated by the examples below.

Schema and configuration requirements for updating child documents use the same Field Storage requirements for atomic updates mentioned above.

Under the hood, Solr conceptually behaves similarly for nested documents as for non-nested documents, it’s just that it applies to entire trees (from the root) of nested documents instead of stand-alone documents. You can expect more overhead because of this. In-place updates avoid that.

Routing Updates using child document Ids in SolrCloud

When SolrCloud receives document updates, the document routing rules for the collection is used to determine which shard should process the update based on the id of the document.

When sending an update that specifies the id of a child document this will not work by default: the correct shard to send the document to is based on the id of the "Root" document for the block the child document is in, not the id of the child document being updated.

Solr offers two solutions to address this:

Clients may specify a _route_ parameter, with the id of the Root document as the parameter value, on each update to tell Solr which shard should process the update.
Clients can use the (default) compositeId router’s "prefix routing" feature when indexing all documents to ensure that all child/descendent documents in a Block use the same id prefix as the Root level document. This will cause Solr’s default routing logic to automatically send child document updates to the correct shard.

Furthermore, you should (sometimes must) specify the Root document’s ID in the _root_ field of this partial update. This is how Solr understands that you are updating a child document, and not a Root document. Without it, Solr only guesses that the _route_ parameter is equivalent, but it may be absent or not equivalent (e.g., when using the implicit router).

All of the examples below use id prefixes, so no _route_ parameter will be necessary for these examples.

For the upcoming examples, we’ll assume an index containing the same documents covered in Indexing Nested Documents:

[{ "id": "P11!prod",
   "name_s": "Swingline Stapler",
   "description_t": "The Cadillac of office staplers ...",
   "skus": [ { "id": "P11!S21",
               "color_s": "RED",
               "price_i": 42,
               "manuals": [ { "id": "P11!D41",
                              "name_s": "Red Swingline Brochure",
                              "pages_i":1,
                              "content_t": "..."
                            } ]
             },
             { "id": "P11!S31",
               "color_s": "BLACK",
               "price_i": 3
             } ],
   "manuals": [ { "id": "P11!D51",
                  "name_s": "Quick Reference Guide",
                  "pages_i":1,
                  "content_t": "How to use your stapler ..."
                },
                { "id": "P11!D61",
                  "name_s": "Warranty Details",
                  "pages_i":42,
                  "content_t": "... lifetime guarantee ..."
                } ]
 },
 { "id": "P22!prod",
   "name_s": "Mont Blanc Fountain Pen",
   "description_t": "A Premium Writing Instrument ...",
   "skus": [ { "id": "P22!S22",
               "color_s": "RED",
               "price_i": 89,
               "manuals": [ { "id": "P22!D42",
                              "name_s": "Red Mont Blanc Brochure",
                              "pages_i":1,
                              "content_t": "..."
                            } ]
             },
             { "id": "P22!S32",
               "color_s": "BLACK",
               "price_i": 67
             } ],
   "manuals": [ { "id": "P22!D52",
                  "name_s": "How To Use A Pen",
                  "pages_i":42,
                  "content_t": "Start by removing the cap ..."
                } ]
 } ]

Modifying Child Document Fields

All of the Atomic Update operations mentioned above are supported for "real" fields of Child Documents:

curl -X POST 'http://localhost:8983/solr/gettingstarted/update?commit=true' -H 'Content-Type: application/json' --data-binary '[
{
  "id": "P11!S31",
  "_root_": "P11!prod",
  "price_i": { "inc": 73 },
  "color_s": { "set": "GREY" }
} ]'

Replacing All Child Documents

As with normal (multiValued) fields, the set keyword can be used to replace all child documents in a pseudo-field:

curl -X POST 'http://localhost:8983/solr/gettingstarted/update?commit=true' -H 'Content-Type: application/json' --data-binary '[
{
  "id": "P22!S22",
  "_root_": "P22!prod",
  "manuals": { "set": [ { "id": "P22!D77",
                          "name_s": "Why Red Pens Are the Best",
                          "content_t": "... correcting papers ...",
                        },
                        { "id": "P22!D88",
                          "name_s": "How to get Red ink stains out of fabric",
                          "content_t": "... vinegar ...",
                        } ] }

} ]'

Adding a Child Document

As with normal (multiValued) fields, the add keyword can be used to add additional child documents to a pseudo-field:

curl -X POST 'http://localhost:8983/solr/gettingstarted/update?commit=true' -H 'Content-Type: application/json' --data-binary '[
{
  "id": "P11!S21",
  "_root_": "P11!prod",
  "manuals": { "add": { "id": "P11!D99",
                        "name_s": "Why Red Staplers Are the Best",
                        "content_t": "Once upon a time, Mike Judge ...",
                      } }
} ]'

Removing a Child Document

As with normal (multiValued) fields, the remove keyword can be used to remove a child document (by id) from it’s pseudo-field:

curl -X POST 'http://localhost:8983/solr/gettingstarted/update?commit=true' -H 'Content-Type: application/json' --data-binary '[
{
  "id": "P11!S21",
  "_root_": "P11!prod",
  "manuals": { "remove": { "id": "P11!D41" } }
} ]'

In-Place Updates

In-place updates are very similar to atomic updates; in some sense, this is a subset of atomic updates. In regular atomic updates, the entire document is reindexed internally during the application of the update. However, in this approach, only the fields to be updated are affected and the rest of the documents are not reindexed internally. Hence, the efficiency of updating in-place is unaffected by the size of the documents that are updated (i.e., number of fields, size of fields, etc.). Apart from these internal differences in efficiency, there is no functional difference between atomic updates and in-place updates.

An atomic update operation is performed using this In-Place approach only when the fields to be updated meet these three conditions:

are non-indexed (indexed="false"), non-stored (stored="false"), single valued (multiValued="false") numeric docValues (docValues="true") fields;
the _version_ field is also a non-indexed, non-stored single valued docValues field; and,
copy targets of updated fields, if any, are also non-indexed, non-stored single valued numeric docValues fields.

To use in-place updates, add a modifier to the field that needs to be updated. The content can be updated or incrementally increased.

set: Set or replace the field value(s) with the specified value(s). May be specified as a single value.
inc: Increments a numeric value by a specific amount. Must be specified as a single numeric value.

Preventing Atomic Updates That Can’t be Done In-Place

Since it can be tricky to ensure that all of the necessary conditions are satisfied to ensure that an update can be done In-Place, Solr supports a request parameter option named update.partial.requireInPlace. When set to true, an atomic update that can not be done In-Place will fail. Users can specify this option when they would prefer that an update request "fail fast" if it can’t be done In-Place.

In-Place Update Example

If the price and popularity fields are defined in the schema as:

<field name="price" type="float" indexed="false" stored="false" docValues="true"/>

<field name="popularity" type="float" indexed="false" stored="false" docValues="true"/>

If the following document exists in our collection:

{
 "id":"mydoc",
 "price":10,
 "popularity":42,
 "categories":["kids"],
 "promo_ids":["a123x"],
 "tags":["free_to_try","buy_now","clearance","on_sale"]
}

And we apply the following update command:

{
 "id":"mydoc",
 "price":{"set":99},
 "popularity":{"inc":20}
}

The resulting document in our collection will be:

{
 "id":"mydoc",
 "price":99,
 "popularity":62,
 "categories":["kids"],
 "promo_ids":["a123x"],
 "tags":["free_to_try","buy_now","clearance","on_sale"]
}

Optimistic Concurrency

Optimistic Concurrency is a feature of Solr that can be used by client applications which update/replace documents to ensure that the document they are replacing/updating has not been concurrently modified by another client application. This feature works by requiring a _version_ field on all documents in the index, and comparing that to a _version_ specified as part of the update command. By default, Solr’s Schema includes a _version_ field, and this field is automatically added to each new document.

In general, using optimistic concurrency involves the following work flow:

A client reads a document. In Solr, one might retrieve the document with the /get handler to be sure to have the latest version.
A client changes the document locally.
The client resubmits the changed document to Solr, for example, perhaps with the /update handler.
If there is a version conflict (HTTP error code 409), the client starts the process over.

When the client resubmits a changed document to Solr, the _version_ can be included with the update to invoke optimistic concurrency control. Specific semantics are used to define when the document should be updated or when to report a conflict.

If the content in the _version_ field is greater than '1' (i.e., '12345'), then the _version_ in the document must match the _version_ in the index.
If the content in the _version_ field is equal to '1', then the document must simply exist. In this case, no version matching occurs, but if the document does not exist, the updates will be rejected.
If the content in the _version_ field is less than '0' (i.e., '-1'), then the document must not exist. In this case, no version matching occurs, but if the document exists, the updates will be rejected.
If the content in the _version_ field is equal to '0', then it doesn’t matter if the versions match or if the document exists or not. If it exists, it will be overwritten; if it does not exist, it will be added.

When documents are added/updated in batches even a single version conflict may lead to rejecting the entire batch. Use the parameter failOnVersionConflicts=false to avoid failure of the entire batch when version constraints fail for one or more documents in a batch.

If the document being updated does not include the _version_ field, and atomic updates are not being used, the document will be treated by normal Solr rules, which is usually to discard the previous version.

When using Optimistic Concurrency, clients can include an optional versions=true request parameter to indicate that the new versions of the documents being added should be included in the response. This allows clients to immediately know what the _version_ is of every document added without needing to make a redundant /get request.

Following are some examples using versions=true in queries:

$ curl -X POST -H 'Content-Type: application/json' 'http://localhost:8983/solr/techproducts/update?versions=true&omitHeader=true' --data-binary '
[ { "id" : "aaa" },
  { "id" : "bbb" } ]'

{
  "adds":[
    "aaa",1632740120218042368,
    "bbb",1632740120250548224]}

In this example, we have added 2 documents "aaa" and "bbb". Because we added versions=true to the request, the response shows the document version for each document.

$ curl -X POST -H 'Content-Type: application/json' 'http://localhost:8983/solr/techproducts/update?_version_=999999&versions=true&omitHeader=true' --data-binary '
  [{ "id" : "aaa",
     "foo_s" : "update attempt with wrong existing version" }]'

{
  "error":{
    "metadata":[
      "error-class","org.apache.solr.common.SolrException",
      "root-error-class","org.apache.solr.common.SolrException"],
    "msg":"version conflict for aaa expected=999999 actual=1632740120218042368",
    "code":409}}

In this example, we’ve attempted to update document "aaa" but specified the wrong version in the request: version=999999 doesn’t match the document version we just got when we added the document. We get an error in response.

$ curl -X POST -H 'Content-Type: application/json' 'http://localhost:8983/solr/techproducts/update?_version_=1632740120218042368&versions=true&commit=true&omitHeader=true' --data-binary '
[{ "id" : "aaa",
   "foo_s" : "update attempt with correct existing version" }]'

{
  "adds":[
    "aaa",1632740462042284032]}

Now we’ve sent an update with a value for _version_ that matches the value in the index, and it succeeds. Because we included versions=true to the update request, the response includes a different value for the _version_ field.

$ curl -X POST -H 'Content-Type: application/json' 'http://localhost:8983/solr/techproducts/update?&versions=true&commit=true&omitHeader=true' --data-binary '
[{ "id" : "aaa", _version_ : 100,
   "foo_s" : "update attempt with wrong existing version embedded in document" }]'

{
  "error":{
    "metadata":[
      "error-class","org.apache.solr.common.SolrException",
      "root-error-class","org.apache.solr.common.SolrException"],
    "msg":"version conflict for aaa expected=100 actual=1632740462042284032",
    "code":409}}

Now we’ve sent an update with a value for _version_ embedded in the document itself. this request fails because we have specified the wrong version. This is useful when documents are sent in a batch and different _version_ values need to be specified for each doc.

$ curl -X POST -H 'Content-Type: application/json' 'http://localhost:8983/solr/techproducts/update?&versions=true&commit=true&omitHeader=true' --data-binary '
[{ "id" : "aaa", _version_ : 1632740462042284032,
   "foo_s" : "update attempt with correct version embedded in document" }]'

{
  "adds":[
    "aaa",1632741942747987968]}

$ curl 'http://localhost:8983/solr/techproducts/query?q=*:*&fl=id,_version_&omitHeader=true'

{
  "response":{"numFound":3,"start":0,"docs":[
      { "_version_":1632740120250548224,
        "id":"bbb"},
      { "_version_":1632741942747987968,
        "id":"aaa"}]
  }}

Finally, we can issue a query that requests the _version_ field be included in the response, and we can see that for the two documents in our example index.

$ curl -X POST -H 'Content-Type: application/json' 'http://localhost:8983/solr/techproducts/update?versions=true&_version_=-1&failOnVersionConflicts=false&omitHeader=true' --data-binary '
[ { "id" : "aaa" },
  { "id" : "ccc" } ]'

{
  "adds":[
    "ccc",1632740949182382080]}

In this example, we have added 2 documents "aaa" and "ccc". As we have specified the parameter _version_=-1, this request should not add the document with the id aaa because it already exists. The request succeeds & does not throw any error because the failOnVersionConflicts=false parameter is specified. The response shows that only document ccc is added and aaa is silently ignored.

For more information, please also see Yonik Seeley’s presentation on NoSQL features in Solr 4 from Apache Lucene EuroCon 2012.

Document Centric Versioning Constraints

Optimistic Concurrency is extremely powerful, and works very efficiently because it uses an internally assigned, globally unique values for the _version_ field. However, in some situations users may want to configure their own document specific version field, where the version values are assigned on a per-document basis by an external system, and have Solr reject updates that attempt to replace a document with an "older" version. In situations like this the DocBasedVersionConstraintsProcessorFactory can be useful.

The basic usage of DocBasedVersionConstraintsProcessorFactory is to configure it in solrconfig.xml as part of the UpdateRequestProcessorChain and specify the name of your custom versionField in your schema that should be checked when validating updates:

<processor class="solr.DocBasedVersionConstraintsProcessorFactory">
  <str name="versionField">my_version_l</str>
</processor>

Note that versionField is a comma-delimited list of fields to check for version numbers. Once configured, this update processor will reject (HTTP error code 409) any attempt to update an existing document where the value of the my_version_l field in the "new" document is not greater then the value of that field in the existing document.

versionField vs _version_

The _version_ field used by Solr for its normal optimistic concurrency also has important semantics in how updates are distributed to replicas in SolrCloud, and MUST be assigned internally by Solr. Users can not re-purpose that field and specify it as the versionField for use in the DocBasedVersionConstraintsProcessorFactory configuration.

DocBasedVersionConstraintsProcessorFactory supports the following additional configuration parameters, which are all optional:

ignoreOldUpdates

A boolean option which defaults to false. If set to true, the update will be silently ignored (and return a status 200 to the client) instead of rejecting updates where the versionField is too low.

deleteVersionParam

A String parameter that can be specified to indicate that this processor should also inspect Delete By Id commands.

The value of this option should be the name of a request parameter that the processor will consider mandatory for all attempts to Delete By Id, and must be be used by clients to specify a value for the versionField which is greater then the existing value of the document to be deleted.

When using this request parameter, any Delete By Id command with a high enough document version number to succeed will be internally converted into an Add Document command that replaces the existing document with a new one which is empty except for the Unique Key and versionField to keeping a record of the deleted version so future Add Document commands will fail if their "new" version is not high enough.

If versionField is specified as a list, then this parameter too must be specified as a comma-delimited list of the same size so that the parameters correspond with the fields.

supportMissingVersionOnOldDocs

This boolean parameter defaults to false, but if set to true allows any documents written before this feature is enabled, and which are missing the versionField, to be overwritten.

Please consult the DocBasedVersionConstraintsProcessorFactory javadocs and test solrconfig.xml file for additional information and example usages.