hanahmily opened a new pull request, #122:
URL: https://github.com/apache/skywalking-banyandb/pull/122

   This introduced a stress test framework to generate fixed traffic to verify 
several behaviors of the banyand server.
    
   ## Verify the flushing mechanism of tsdb
   
   Most caches of tsdb are based on the badger which limits the memory usage by 
a fixed number. But the inverted index's memtable gets flushed by an event from 
the main storage. 
   
   The below screenshot shows the flushing of the main storage.
   
   ![main storage 
flushing](https://user-images.githubusercontent.com/1960537/170975585-d7a8c68a-016e-4746-a02c-2d4f8c3f436c.png)
   
   The below diagram indicates the inverted indices flushing under the 
controller of the main storage.
   ![inverted index 
flushing](https://user-images.githubusercontent.com/1960537/170975662-ec53e3cb-30da-4944-965f-9b906f21c8e7.png)
   
   ## Check encoding algorithm
   
   This diagram shows the compression ratio of a stream under the case that 
each instance keeps sending the element to the database server. That's why the 
ratio is such high(about 98%). We will leverage the e2e test to get a more 
practical result in the next round. This result only indicates that the 
algorithm is working as expected.
   
   
![image](https://user-images.githubusercontent.com/1960537/170977140-9f3e2d99-a07f-496e-aaba-b32edcc0aade.png)
   
   This is the gorilla encoding algorithm's result which is similar to the 
theory result, around 30%
   
   
![image](https://user-images.githubusercontent.com/1960537/170977904-759afe4c-5945-4607-97eb-f358ecd8a9d3.png)
   
   
   
   Signed-off-by: Gao Hongtao <hanahm...@gmail.com>


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: notifications-unsubscr...@skywalking.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org

Reply via email to