You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ofbiz.apache.org by "Adrian Crum (JIRA)" <ji...@apache.org> on 2015/03/20 11:11:39 UTC

[jira] [Comment Edited] (OFBIZ-5534) Dirty reads from entity cache

    [ https://issues.apache.org/jira/browse/OFBIZ-5534?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14371066#comment-14371066 ] 

Adrian Crum edited comment on OFBIZ-5534 at 3/20/15 10:11 AM:
--------------------------------------------------------------

This issue is now a year old, and I would like to help get it resolved. Until Adam can produce his solution, I have my own that we can consider.

I don't think we need to use XAResource. The problem is fairly easy to solve, so using XAResource will complicate things unnecessarily.

1. We need to add a bit of functionality to the existing transaction code:
  a. Associate org.ofbiz.entity.cache.Cache instances (one for each Delegator instance) to each UserTransaction instance.
  b. Add Publish/Subscribe (listener) functionality to TransactionUtil. Listeners are notified post-commit and post-rollback.

2. Add functionality to the Delegator:
  a. Make each instance a transaction listener.
  b. On post-commit notification, copy transaction cache to internal cache.
  c. On post-rollback notification, clear transaction cache.

3. Modify Delegator cache behavior:
  a. On find methods: If a transaction is in place, check transaction cache, else check internal cache.
  b. If a miss, execute SQL SELECT.
  c. If a transaction is in place, put results in transaction cache, else put results in internal cache.



was (Author: adrianc@hlmksw.com):
This issue is now a year old, and I would like to help get it resolved. Until Adam can produce his solution, I have my own that we can consider.

I don't think we need to use XAResource. The problem is fairly easy to solve, so using XAResource will complicate things unnecessarily.

1. We need to add a bit of functionality to the existing transaction code:
  a. Associate org.ofbiz.entity.cache.Cache instances (one for each Delegator instance) to each UserTransaction instance.
  b. Add Publish/Subscribe (listener) functionality to TransactionUtil. Listeners are notified post-commit and post-rollback.

2. Add functionality to the Delegator:
  a. Make each instance a transaction listener.
  b. On post-commit notification, copy transaction cache to internal cache.
  c. On post-rollback notification, clear transaction cache.

3. Modify Delegator cache behavior:
  a. On find methods, check its internal cache first.
  b. If a miss and transaction is in place, check transaction cache.
  c. If a miss, execute SQL SELECT.
  d. If a transaction is in place, put results in transaction cache, else put results in internal cache.


> Dirty reads from entity cache
> -----------------------------
>
>                 Key: OFBIZ-5534
>                 URL: https://issues.apache.org/jira/browse/OFBIZ-5534
>             Project: OFBiz
>          Issue Type: Bug
>          Components: framework
>    Affects Versions: Release Branch 11.04, Release Branch 12.04, Release Branch 13.07, Trunk
>            Reporter: Jacopo Cappellato
>            Assignee: Adam Heath
>         Attachments: unittest-dirtyreadsfromcache.patch
>
>
> Even if database transaction isolation level is set to "ReadCommitted", a transaction can get uncommitted records from the entity cache, under certain circumstances.
> Here is a test case:
> 1) transaction T1 creates a record
> 2) T1 calls findOne to retrieve it: the read will be successful (because T1 can read data generated in T1 even if not committed) and the record will be added to the cache
> 3) T1 calls another service with requireNewTransaction set to true: this will run the second service in a new transaction T2
> 4) T2 calls findOne to retrieve the record created by T1: if useCache is set to false it should fail; if useCache is set to true it will succeed (getting the record from cache, even if the data is not committed to the db)
> The result in #4 is wrong because T2 should not see the uncommitted record
> We should never add uncommitted records to the entity cache.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)