Rework data migration docs and add reindex to Serverless steps#4914
Rework data migration docs and add reindex to Serverless steps#4914kilfoyle wants to merge 15 commits intoelastic:mainfrom
Conversation
✅ Vale Linting ResultsNo issues found on modified lines! The Vale linter checks documentation changes against the Elastic Docs style guide. To use Vale locally or report issues, refer to Elastic style guide for Vale. |
PeteGillinElastic
left a comment
There was a problem hiding this comment.
Thanks! A few thoughts below.
|
|
||
| :::{important} | ||
| Kibana assets must be migrated separately using the {{kib}} [export/import APIs](https://www.elastic.co/docs/api/doc/kibana/group/endpoint-saved-objects) or recreated manually. | ||
| Templates, data stream definitions, and ILM policies, must be in place _before_ you start data migration. |
There was a problem hiding this comment.
There's an important wrinkle here. You do want your templates set up with things like the correct index mappings and most other index settings. But a lot of the time you do not want your ingest pipelines in place before you reindex. The reason being that the data will have been run through whatever pipeline is in place in ECH when it was ingested there, and (unless it happens to be idempotent) you don't want to run it through the same pipeline a second time in serverless.
This is potentially an awkward thing from an ergonomics perspective. I think that @pete-naylor has thought more about how to handle it smoothly than I have, so I'll defer to him.
There was a problem hiding this comment.
Of course, if you are going to be switching the thing that was previously writing into ECH to write into serverless instead, you presumably do want the ingest pipelines in place before you start doing that...
There was a problem hiding this comment.
Thanks for catching that @PeteGillinElastic! I rephrased the warning to try and catch the nuance about ingest pipelines.
Not really related to this, but I've also expanded the parent Migrate your Elasticsearch data landing page to summarize the different data types and migration methods, so maybe that'll help.
|
@shainaraskas and @eedugon Thanks very much for your feedback! Edu, the detail you provided about the various data types is super helpful so I've incorporated that into the main Migrate your Elasticsearch data page. |
Wow, the new content feels really awesome, that's exactly what was missing! It would be great if we get some review by product experts to ensure everything is technically accurate. I'll review and share some comments soon. |
|
Here's a suggestion for an alternate format:
For example, for the "To Serverless" scenario, currently the only supported From for reindex from remote is Hosted:
*Technical Preview |
|
@ppf2 I can try what you suggest. To start, I'd need to verify that the tables are correct and exhaustive. Would you be able confirm if these are right? Or would you have an idea who best to ask? I do worry that these tables would over-complicate things a bit. The reason for the current page organization is that it imitates how we organize the "Deploy" section, but I'm open to changing that. To Elastic Cloud Hosted:
To Elastic Cloud Enterprise:
To Elastic Cloud on Kubernetes:
To Elastic Cloud Serverless:
To Elastic Self-managed:
|
|
@ppf2 I've opened a new PR that follows your suggestion of having a single page for each migration method: Rework migrate data docs (Take 2) I think it could work nicely. |
|
@kilfoyle : the serverless row in one of these blocks is probably incorrect:
Whatever is the right value should probably be the same in both cases. And in this block I have a question:
How is it possible to support using Logstash to send data from ECE or ECK to Serverless but not from a self-managed cluster to Serverless? I'm tempted to say that has to be a mistake, as technically speaking, an ECK or ECE Elasticsearch cluster is the same as a self-managed cluster: they run on self-managed infrastructure, and all of them have HTTP endpoints controlled and created by the user. |
eedugon
left a comment
There was a problem hiding this comment.
small comment added, sorry I haven't been able to review the entire PR again. If needed please let me know!! :)
|
Thanks @edu! I've added your suggestions in the new PR. I'll close this PR in favour of the new one: #5244 |


Please see the newer version of this PR: Rework migrate data docs (Take 2)
This PR:
@PeteGillinElastic @pete-naylor I'd appreciate if either of you can review the reindex from remote examples. Please let me know if I've missed anything.
Closes: #4353
Rel: https://github.com/elastic/docs-team/issues/30
Rel: https://github.com/elastic/docs-content-internal/issues/467