You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@bookkeeper.apache.org by GitBox <gi...@apache.org> on 2018/02/24 01:42:35 UTC

[GitHub] sijie commented on issue #570: Multiple active entrylogs

sijie commented on issue #570: Multiple active entrylogs
URL: https://github.com/apache/bookkeeper/issues/570#issuecomment-368186471
 
 
   ```
    The implementation would become simpler, because it is a complex logic to handle and maintain slotmap in configured number of Entrylogs per LedgerDir design/implementation and it needs to be thoroughly analyzed/tested from multi-thread perspective. But most importantly the garbagecollection component becomes much simpler because there is no need of compaction in entrylog per ledger approach. So there is huge gain in indirect performance because of absence of IO activity for compaction and huge improvement in space reclamation during compaction.
   ```
   
   This needs to be clarified. It would have the benefits if you have small number of ledgers, I don't think it is good for large number of ledgers. It is good if you think from Salesforce's use case; however it is not, if you think from other use cases. This is good to be called out before you are claiming all the benefits here.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services