You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I've been trying to learn Faster for almost a week now and facing some difficulties.
I wanted to see how FasterKV store handles Concurrency. After viewing a few samples it was evident that we can have multiple producer threads with dedicated sessions for dedicated stores and then a different set of consumer threads that can read from the same store either concurrently or via recovery. This architecture leads to sizing each index and log separately, which is fine in some cases.
Next, I wanted to see Concurrent Producer Consumers for the same store; this led me to the SumStore example inside the Playground folder. I failed to grasp why a few entries have different values for the Expected & Found number-of-clicks reported in the last Asset for both Concurrent & Recovery tests.
Concurrency Test
Debug error for AdId 14099: Expected (2), Found(1)
Debug error for AdId 15069: Expected (2), Found(1)
Recovery Test
Debug error for AdId 5535694: Expected (106), Found(105)
Debug error for AdId 6441859: Expected (104), Found(103)
Does it have to do anything with a possible Race Condition between two threads when on the InitialUpdater phase of RMW, i.e., when the key is not present in the index?
I have tried calling Refresh and CompletePending(true) but they did not help. Please help.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hi, I've been trying to learn Faster for almost a week now and facing some difficulties.
I wanted to see how FasterKV store handles Concurrency. After viewing a few samples it was evident that we can have multiple producer threads with dedicated sessions for dedicated stores and then a different set of consumer threads that can read from the same store either concurrently or via recovery. This architecture leads to sizing each index and log separately, which is fine in some cases.
Next, I wanted to see Concurrent Producer Consumers for the same store; this led me to the SumStore example inside the Playground folder. I failed to grasp why a few entries have different values for the Expected & Found number-of-clicks reported in the last Asset for both Concurrent & Recovery tests.
Concurrency Test
Debug error for AdId 14099: Expected (2), Found(1)
Debug error for AdId 15069: Expected (2), Found(1)
Recovery Test
Debug error for AdId 5535694: Expected (106), Found(105)
Debug error for AdId 6441859: Expected (104), Found(103)
Does it have to do anything with a possible Race Condition between two threads when on the InitialUpdater phase of RMW, i.e., when the key is not present in the index?
I have tried calling Refresh and CompletePending(true) but they did not help. Please help.
Beta Was this translation helpful? Give feedback.
All reactions