Troubleshooting Elasticsearch (PREMIUM SELF)

Use the following information to troubleshoot Elasticsearch issues.

Set configurations in the Rails console

See Starting a Rails console session.

List attributes

To list all available attributes:

  1. Open the Rails console (gitlab rails c).
  2. Run the following command:

The output contains all the settings available in Elasticsearch integration, such as elasticsearch_indexing, elasticsearch_url, elasticsearch_replicas, and elasticsearch_pause_indexing.

Set attributes

To set an Elasticsearch integration setting, run a command like:

ApplicationSetting.last.update(elasticsearch_url: '<your ES URL and port>')


ApplicationSetting.last.update(elasticsearch_indexing: false)

Get attributes

To check if the settings have been set in Elasticsearch integration or in the Rails console, run a command like:




Change the password

To change the Elasticsearch password, run the following commands:

es_url = Gitlab::CurrentSettings.current_application_settings

# Confirm the current Elasticsearch URL

# Set the Elasticsearch URL
es_url.elasticsearch_url = "http://<username>:<password><port>"

# Save the change!

View logs

One of the most valuable tools for identifying issues with the Elasticsearch integration are logs. The most relevant logs for this integration are:

  1. sidekiq.log - All of the indexing happens in Sidekiq, so much of the relevant logs for the Elasticsearch integration can be found in this file.
  2. elasticsearch.log - There are additional logs specific to Elasticsearch that are sent to this file that may contain useful diagnostic information about searching, indexing or migrations.

Here are some common pitfalls and how to overcome them.

Common terminology

  • Lucene: A full-text search library written in Java.
  • Near real time (NRT): Refers to the slight latency from the time to index a document to the time when it becomes searchable.
  • Cluster: A collection of one or more nodes that work together to hold all the data, providing indexing and search capabilities.
  • Node: A single server that works as part of a cluster.
  • Index: A collection of documents that have somewhat similar characteristics.
  • Document: A basic unit of information that can be indexed.
  • Shards: Fully-functional and independent subdivisions of indices. Each shard is actually a Lucene index.
  • Replicas: Failover mechanisms that duplicate indices.

How can you verify that your GitLab instance is using Elasticsearch?

There are a couple of ways to achieve that:

  • When you perform a search, in the upper-right corner of the search results page, Advanced search is enabled is displayed. This is always correctly identifying whether the current project/namespace being searched is using Elasticsearch.

  • From the Admin Area under Settings > Advanced Search check that the advanced search settings are checked.

    Those same settings there can be obtained from the Rails console if necessary:

    ::Gitlab::CurrentSettings.elasticsearch_search?         # Whether or not searches will use Elasticsearch
    ::Gitlab::CurrentSettings.elasticsearch_indexing?       # Whether or not content will be indexed in Elasticsearch
    ::Gitlab::CurrentSettings.elasticsearch_limit_indexing? # Whether or not Elasticsearch is limited only to certain projects/namespaces
  • Confirm searches use Elasticsearch by accessing the rails console and running the following commands:

    u = User.find_by_email('email_of_user_doing_search')
    s =, {:search => 'search_term'})
    pp s.search_objects.class

    The output from the last command is the key here. If it shows:

    • ActiveRecord::Relation, it is not using Elasticsearch.
    • Kaminari::PaginatableArray, it is using Elasticsearch.
  • If Elasticsearch is limited to specific namespaces and you need to know if Elasticsearch is being used for a specific project or namespace, you can use the Rails console:

    ::Gitlab::CurrentSettings.search_using_elasticsearch?(scope: Namespace.find_by_full_path("/my-namespace"))
    ::Gitlab::CurrentSettings.search_using_elasticsearch?(scope: Project.find_by_full_path("/my-namespace/my-project"))

Troubleshooting indexing

Troubleshooting indexing issues can be tricky. It can pretty quickly go to either GitLab support or your Elasticsearch administrator.

The best place to start is to determine if the issue is with creating an empty index. If it is, check on the Elasticsearch side to determine if the gitlab-production (the name for the GitLab index) exists. If it exists, manually delete it on the Elasticsearch side and attempt to recreate it from the recreate_index Rake task.

If you still encounter issues, try creating an index manually on the Elasticsearch instance. The details of the index aren't important here, as we want to test if indices can be made. If the indices:

  • Cannot be made, speak with your Elasticsearch administrator.
  • Can be made, Escalate this to GitLab support.

If the issue is not with creating an empty index, the next step is to check for errors during the indexing of projects. If errors do occur, they stem from either the indexing:

  • On the GitLab side. You need to rectify those. If they are not something you are familiar with, contact GitLab support for guidance.
  • Within the Elasticsearch instance itself. See if the error is documented and has a fix. If not, speak with your Elasticsearch administrator.

If the indexing process does not present errors, check the status of the indexed projects. You can do this via the following Rake tasks:


  • Everything is showing at 100%, escalate to GitLab support. This could be a potential bug/issue.
  • You do see something not at 100%, attempt to reindex that project. To do this, run sudo gitlab-rake gitlab:elastic:index_projects ID_FROM=<project ID> ID_TO=<project ID>.

If reindexing the project shows:

  • Errors on the GitLab side, escalate those to GitLab support.
  • Elasticsearch errors or doesn't present any errors at all, reach out to your Elasticsearch administrator to check the instance.

You updated GitLab and now you can't find anything

We continuously make updates to our indexing strategies and aim to support newer versions of Elasticsearch. When indexing changes are made, it may be necessary for you to reindex after updating GitLab.

You indexed all the repositories but you can't get any hits for your search term in the UI

Make sure you indexed all the database data.

If there aren't any results (hits) in the UI search, check if you are seeing the same results via the rails console (sudo gitlab-rails console):

u = User.find_by_username('your-username')
s =, {:search => 'search_term', :scope => 'blobs'})
pp s.search_objects.to_a

Beyond that, check via the Elasticsearch Search API to see if the data shows up on the Elasticsearch side:

curl --request GET <elasticsearch_server_ip>:9200/gitlab-production/_search?q=<search_term>

More complex Elasticsearch API calls are also possible.

If the results:

NOTE: The above instructions are not to be used for scenarios that only index a subset of namespaces.

See Elasticsearch Index Scopes for more information on searching for specific types of data.

You indexed all the repositories but then switched Elasticsearch servers and now you can't find anything

You must re-run all the Rake tasks to reindex the database, repositories, and wikis.

The indexing process is taking a very long time

The more data present in your GitLab instance, the longer the indexing process takes.

There are some projects that weren't indexed, but you don't know which ones

You can run sudo gitlab-rake gitlab:elastic:projects_not_indexed to display projects that aren't indexed.

No new data is added to the Elasticsearch index when you push code

NOTE: This was fixed in GitLab 13.2 and the Rake task is not available for versions greater than that.

When performing the initial indexing of blobs, we lock all projects until the project finishes indexing. It could happen that an error during the process causes one or multiple projects to remain locked. To unlock them, run:

sudo gitlab-rake gitlab:elastic:clear_locked_projects

Indexing fails with error: elastic: Error 429 (Too Many Requests)

If ElasticCommitIndexerWorker Sidekiq workers are failing with this error during indexing, it usually means that Elasticsearch is unable to keep up with the concurrency of indexing request. To address change the following settings:

Indexing is very slow or fails with rejected execution of coordinating operation messages

Bulk requests getting rejected by the Elasticsearch nodes are likely due to load and lack of available memory. Ensure that your Elasticsearch cluster meets the system requirements and has enough resources to perform bulk operations. See also the error "429 (Too Many Requests)".

Last resort to recreate an index

There may be cases where somehow data never got indexed and it's not in the queue, or the index is somehow in a state where migrations just cannot proceed. It is always best to try to troubleshoot the root cause of the problem by viewing the logs.

If there are no other options, then you always have the option of recreating the entire index from scratch. If you have a small GitLab installation, this can sometimes be a quick way to resolve a problem, but if you have a large GitLab installation, then this might take a very long time to complete. Until the index is fully recreated, your index does not serve correct search results, so you may want to disable Search with Elasticsearch while it is running.

If you are sure you've read the above caveats and want to proceed, then you should run the following Rake task to recreate the entire index from scratch:

For Omnibus installations

sudo gitlab-rake gitlab:elastic:index

For installations from source

cd /home/git/gitlab
sudo -u git -H bundle exec rake gitlab:elastic:index

Troubleshooting performance

Troubleshooting performance can be difficult on Elasticsearch. There is a ton of tuning that can be done, but the majority of this falls on shoulders of a skilled Elasticsearch administrator.

Generally speaking, ensure:

  • The Elasticsearch server is not running on the same node as GitLab.
  • The Elasticsearch server have enough RAM and CPU cores.
  • That sharding is being used.

Going into some more detail here, if Elasticsearch is running on the same server as GitLab, resource contention is very likely to occur. Ideally, Elasticsearch, which requires ample resources, should be running on its own server (maybe coupled with Logstash and Kibana).

When it comes to Elasticsearch, RAM is the key resource. Elasticsearch themselves recommend:

  • At least 8 GB of RAM for a non-production instance.
  • At least 16 GB of RAM for a production instance.
  • Ideally, 64 GB of RAM.

For CPU, Elasticsearch recommends at least 2 CPU cores, but Elasticsearch states common setups use up to 8 cores. For more details on server specs, check out the Elasticsearch hardware guide.

Beyond the obvious, sharding comes into play. Sharding is a core part of Elasticsearch. It allows for horizontal scaling of indices, which is helpful when you are dealing with a large amount of data.

With the way GitLab does indexing, there is a huge amount of documents being indexed. By using sharding, you can speed up the ability of Elasticsearch to locate data because each shard is a Lucene index.

If you are not using sharding, you are likely to hit issues when you start using Elasticsearch in a production environment.

An index with only one shard has no scale factor and is likely to encounter issues when called upon with some frequency. See the Elasticsearch documentation on capacity planning.

The easiest way to determine if sharding is in use is to check the output of the Elasticsearch Health API:

  • Red means the cluster is down.
  • Yellow means it is up with no sharding/replication.
  • Green means it is healthy (up, sharding, replicating).

For production use, it should always be green.

Beyond these steps, you get into some of the more complicated things to check, such as merges and caching. These can get complicated and it takes some time to learn them, so it is best to escalate/pair with an Elasticsearch expert if you need to dig further into these.

Feel free to reach out to GitLab support, but this is likely to be something a skilled Elasticsearch administrator has more experience with.

Issues with migrations

Please ensure you've read about Elasticsearch Migrations.

If there is a halted migration and your elasticsearch.log file contain errors, this could potentially be a bug/issue. Escalate to GitLab support if retrying migrations does not succeed.

Can't specify parent if no parent field has been configured error

If you enabled Elasticsearch before GitLab 8.12 and have not rebuilt indices, you get exceptions in lots of different cases:

Elasticsearch::Transport::Transport::Errors::BadRequest([400] {
    "error": {
        "root_cause": [{
            "type": "illegal_argument_exception",
            "reason": "Can't specify parent if no parent field has been configured"
        "type": "illegal_argument_exception",
        "reason": "Can't specify parent if no parent field has been configured"
    "status": 400

This is because we changed the index mapping in GitLab 8.12 and the old indices should be removed and built from scratch again, see details in the update guide.


If you have this exception (just like in the case above but the actual message is different) please check if you have the correct Elasticsearch version and you met the other requirements. There is also an easy way to check it automatically with sudo gitlab-rake gitlab:check command.


[413] {"Message":"Request size exceeded 10485760 bytes"}

This exception is seen when your Elasticsearch cluster is configured to reject requests above a certain size (10MiB in this case). This corresponds to the http.max_content_length setting in elasticsearch.yml. Increase it to a larger size and restart your Elasticsearch cluster.

AWS has fixed limits for this setting ("Maximum size of HTTP request payloads"), based on the size of the underlying instance.

Faraday::TimeoutError (execution expired) error when using a proxy

Set a custom gitlab_rails['env'] environment variable, called no_proxy with the IP address of your Elasticsearch host.

My single node Elasticsearch cluster status never goes from yellow to green even though everything seems to be running properly

For a single node Elasticsearch cluster the functional cluster health status is yellow (never green) because the primary shard is allocated but replicas cannot be as there is no other node to which Elasticsearch can assign a replica. This also applies if you are using the Amazon OpenSearch service.

WARNING: Setting the number of replicas to 0 is discouraged (this is not allowed in the GitLab Elasticsearch Integration menu). If you are planning to add more Elasticsearch nodes (for a total of more than 1 Elasticsearch) the number of replicas needs to be set to an integer value larger than 0. Failure to do so results in lack of redundancy (losing one node corrupts the index).

If you have a hard requirement to have a green status for your single node Elasticsearch cluster, please make sure you understand the risks outlined in the previous paragraph and then run the following query to set the number of replicas to 0(the cluster no longer tries to create any shard replicas):

curl --request PUT localhost:9200/gitlab-production/_settings --header 'Content-Type: application/json' \
     --data '{
       "index" : {
         "number_of_replicas" : 0

health check timeout: no Elasticsearch node available error in Sidekiq

If you're getting a health check timeout: no Elasticsearch node available error in Sidekiq during the indexing process:

Gitlab::Elastic::Indexer::Error: time="2020-01-23T09:13:00Z" level=fatal msg="health check timeout: no Elasticsearch node available"

You probably have not used either http:// or https:// as part of your value in the "URL" field of the Elasticsearch Integration Menu. Please make sure you are using either http:// or https:// in this field as the Elasticsearch client for Go that we are using needs the prefix for the URL to be accepted as valid. After you have corrected the formatting of the URL, delete the index (via the dedicated Rake task) and reindex the content of your instance.

My Elasticsearch cluster has a plugin and the integration is not working

Certain 3rd party plugins may introduce bugs in your cluster or for whatever reason may be incompatible with our integration. You should try disabling plugins so you can rule out the possibility that the plugin is causing the problem.

Elasticsearch code_analyzer doesn't account for all code cases

The code_analyzer pattern and filter configuration is being evaluated for improvement. We have fixed most edge cases that were not returning expected search results due to our pattern and filter configuration.

Improvements to the code_analyzer pattern and filters are being discussed in epic 3621.

Some binary files may not be searchable by name

In GitLab 13.9, a change was made where binary file names are being indexed. However, without indexing all projects' data from scratch, only binary files that are added or updated after the GitLab 13.9 release are searchable.

How does advanced search handle private projects?

Advanced search stores all the projects in the same Elasticsearch indices, however, searches only surface results that can be viewed by the user. Advanced search honors all permission checks in the application by filtering out projects that a user does not have access to at search time.

Role mapping when using fine-grained access control with AWS Elasticsearch or OpenSearch

When using fine-grained access control with an IAM role or a role created using OpenSearch Dashboards, you might encounter the following error:

{"error":{"root_cause":[{"type":"security_exception","reason":"no permissions for [indices:data/write/bulk] and User [name=arn:aws:iam::xxx:role/INSERT_ROLE_NAME_HERE, backend_roles=[arn:aws:iam::xxx:role/INSERT_ROLE_NAME_HERE], requestedTenant=null]"}],"type":"security_exception","reason":"no permissions for [indices:data/write/bulk] and User [name=arn:aws:iam::xxx:role/INSERT_ROLE_NAME_HERE, backend_roles=[arn:aws:iam::xxx:role/INSERT_ROLE_NAME_HERE], requestedTenant=null]"},"status":403}

To fix this, you need to map the roles to users in Kibana.