#community-help

Switching from Elastic to Typesense on k8s Cluster

TLDR Nicolo needed to know capabilities and features of Typesense. Jason and William provided explanations and shared experiences using it along with documentation links. Nicolo appreciated the insights.

Powered by Struct AI

2

1

Mar 10, 2022 (22 months ago)
Nicolo
Photo of md5-009fb5f5f979eef4f6f29276c877e1e0
Nicolo
11:07 PM
Hi, there is a question that I have but I cannot find a solution to it.
Right now I'm using elastic but is driving me nuts with costs, and a lot of issues on my k8s cluster.
I only have 1k documents of 3 different kind of collections.

I was wondering if:
• I can search on multiple collections at the same time.
• the collections and the pushed documents are persitent? (on container volume or something)
◦ if not Is there some sort of backend that will be implemented in future like mongo?
• I can scale horizontally or there are some limitaions?
Jason
Photo of md5-8813087cccc512313602b6d9f9ece19f
Jason
11:38 PM
> I can search on multiple collections at the same time.
Yup. It's called federated search / multi search in Typesense: https://typesense.org/docs/0.22.2/api/documents.html#federated-multi-search

> the collections and the pushed documents are persitent?
They are, as long as you make sure the underlying disk / volume is persistent.
11:39
Jason
11:39 PM
> I can scale horizontally or there are some limitaions?
You can setup a cluster with 3 or 5 nodes and use that as a form of horizontal read scaling, since data is replicated to all those nodes.
11:40
Jason
11:40 PM
Technically you can add an odd number of nodes, but with every node you add write latency is takes a slight hit since data has to be acknowledged by a majority of the nodes
Mar 11, 2022 (22 months ago)
William
Photo of md5-94107f2c428564cb51e7a3187ed1dc18
William
04:32 AM
We’re searching across ~10 collections at a time, and it’s super awesome. Federated search works really well. Our use case is a CRM so we have many doc types (invoices, customers, notes, etc) and it’s magical to let them instant-search everything in a single unified search screen.

One thing - you may be surprised how well a small instance performs.. we now use Typesense Cloud, but when we ran things on-prem, we were absolutely hammering a tiny typesense instance and I don’t think we ever managed to really bog it down enough to need much scaling.

2

Nicolo
Photo of md5-009fb5f5f979eef4f6f29276c877e1e0
Nicolo
07:31 PM
Thanks for the answers. I'll take a look on monday to see if I can make it run on my k8s cluster correctly.
Any docs on how scale it horizontally? I suppose that is peering what you mean with scaling... is that right?

BTW I've seen tha tthe docker image dont have a volume specified on the dockerfile!
Jason
Photo of md5-8813087cccc512313602b6d9f9ece19f
Jason
07:41 PM
> I suppose that is peering what you mean with scaling... is that right?
Correct. Here are the docs for this: https://typesense.org/docs/guide/high-availability.html

> BTW I've seen tha tthe docker image dont have a volume specified on the dockerfile!
You will have to mount an external volume and get Typesense to use it like this: https://typesense.org/docs/guide/install-typesense.html#from-the-docker-image
07:42
Jason
07:42 PM
Nicolo
Photo of md5-009fb5f5f979eef4f6f29276c877e1e0
Nicolo
07:56 PM
thanks a lot for the infos! I'm really looking forward to move away from elastic.

1

Typesense

Lightning-fast, open source search engine for everyone | Knowledge Base powered by Struct.AI

Indexed 3015 threads (79% resolved)

Join Our Community

Similar Threads

Understanding Typesense's Functionality and Usage as a Search Engine

Tim asked about Typesense's function as a DB storage. Jason clarified Typesense functions primarily as a search engine and is not meant to be a primary data store. They also discussed data management strategies using Typesense. Joe provided further insight and recommended resources.

7
33mo

Troubleshooting Typesense Document Import Error

Christopher had trouble importing 2.1M documents into Typesense due to memory errors. Jason clarified the system requirements, explaining the correlation between RAM and dataset size, and ways to tackle the issue. They both also discussed database-like query options.

3

30
10mo

Optimizing Typesense Implementation for Large Collections

Oskar faced performance issues with his document collection in Typesense due to filter additions. Jason suggested trying a newer Typesense build and potentially partitioning the data into country-wise collections. They also discussed reducing network latency with CDN solutions.

5

67
11mo

Discussing Multi-tenant Indexing and Typesense Clusters

Semyon enquired about search and indexing performance for multi-tenant indexing. Virginio suggested using Typesense cloud, but Semyon noted the cost difference against bare metal instances. Jason proposed vertical scaling before multi-cluster setup.

3

6
5mo

Addressing Typesense Server Issues and Optimization Needs

Robert had an issue with a 'stuck' typesense server. Jason and Kishore Nallan gave advice on handling writes, configuration for high search volumes, and running multiple typesense instances. They also recommended monitoring CPU usage and updating the server version for bug fixes.

1

30
14mo