Hello,
Today I did performance tests (some kind of) between Linode 4GB and AWS
t2.medium - just to see which one is faster in this situation.
Test results are quite similar between these two - which is a bit
interesting (or probably not - let me know what do you think), because AWS
t2.medium are 2 vCPU, but Linode 4GB are 4vCPU and SSD.
Any additional information you can find in spreadsheet.
Test methodology - each request/query was executed in sync order 100 times
(so 1 requests isn't cached - all others I assume is)
Data was indexed in AWS machine and after that synced (with rsync) to
Linode (so data set is completely same and both instances Elasticsearch
config is completely same).
- Java version same (Open-jdk 7)
- Linux dist. same (Ubuntu 14.04 64bit)
- 1 index/1 shard
- Custom tokenization process (which generates a lot of tokens when
indexing happens - that's why there is 8.4GB in total from 3.2GB of RAW
data)
As measurement I take "took" property from Elasticsearch response.
This wasn't stress test or something like that - each query was executed
100 times in sync order - just to if there is difference.
Cluster (single node) - wasn't on stress or something like this - just
these queries...
Spreadsheet
- https://docs.google.com/spreadsheets/d/1Zoxtww3Vo1EFTzwVp08eHK7i8lHriszPNH31EiUZcAc/edit?usp=sharing
So, the question is - Why AWS results are almost same as Linode? I would
expect that Linode is faster than AWS t2.medium.
Thank you!
--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/8758afcd-232f-43e2-a3ef-638df27d2fbe%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.