#community-help

Discussing Synchronization Delays in SDN across Multiple Regions

TLDR Patrik inquired about synchronization speed in SDN usage across regions. Kishore Nallan explained that data replication happens asynchronously, with latency dependent on geographical spread and node locations.

12mo

Join the chat

Dec 03, 2022 (12 months ago)

Photo of md5-1e5d7ffd1cc67805b3b347e7edaa1b1d

Patrik

10:06 AM

hello, how fast is synchronization when SDN in multiple reagions used? there are consistent responses in all regions, or there can be different responses for some time? thanks

Photo of md5-4e872368b2b2668460205b409e95c2ea

Kishore Nallan

10:07 AM

The replication of data happens asynchronously. So there can be a few seconds of lag between the instances across regions. Actual latency will depend on how far out the nodes are spread, and who is the elected leader etc.

10:09

Kishore Nallan

10:09 AM

For e.g. a round trip time between Cape Town to Tokyo us ~400 ms.

Patrik

10:10 AM

i see but its not great for administrations.. example, admin add post and should see this post in list, where there is lag few seconds and this post wont be there it will be weird :thinking_face:

Kishore Nallan

10:11 AM

Few seconds is the worst case, usually it shows up under 800 ms depending on the geographical spread.

10:13

Kishore Nallan

10:13 AM

But I know what you mean. Synchronous replication might be a mode we could support in future.

Typesense

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

Indexed 3005 threads (79% resolved)

Join Our Community

Similar Threads

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.

11mo

Geo Load-Balanced Endpoint Causing Slow Performance

Edward reported slow responses in their geo load-balanced setup. Kishore Nallan suggested eliminating search_time_ms from the test, which revealed that the issue was not with the LB endpoint. The team will work on improving query latency with facets.

9mo

Discussion on Performance and Scalability for Multiple Term Search

Bill asks the best way for multi-term searches in a recommendation system they developed. Kishore Nallan suggested using embeddings and remote embedder or storing and averaging vectors. Despite testing several suggested solutions, Bill continued to face performance issues, leading to unresolved discussions about scalability and recommendation system performance.

105