Re: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Chetan Mehrotra
Created OAK-1294 to track this Chetan Mehrotra [1] https://issues.apache.org/jira/browse/OAK-1294 On Wed, Dec 18, 2013 at 3:15 PM, Chetan Mehrotra wrote: > The patch works great!! > > With patch applied I get similar timings as I got by disabling session > refresh in LockOperation. > Chetan Mehr

Re: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Chetan Mehrotra
The patch works great!! With patch applied I get similar timings as I got by disabling session refresh in LockOperation. Chetan Mehrotra On Wed, Dec 18, 2013 at 2:47 PM, Michael Dürig wrote: > > > On 18.12.13 10:12 , Marcel Reutegger wrote: >>> >>> We could rebase the branch and then rebase the

Re: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Michael Dürig
On 18.12.13 10:12 , Marcel Reutegger wrote: We could rebase the branch and then rebase the in memory changes on top of the rebased branch. This would get us rid of the branch commit. But wouldn't get us rid of the rebase operation on the persisted branch. So I'm not too sure this will gain us a

RE: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Marcel Reutegger
> We could rebase the branch and then rebase the in memory changes on top > of the rebased branch. This would get us rid of the branch commit. But > wouldn't get us rid of the rebase operation on the persisted branch. So > I'm not too sure this will gain us a lot. hmm, you are right. this may stil

Re: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Michael Dürig
On 18.12.13 9:20 , Marcel Reutegger wrote: Hi, Purging (should) only happen for refresh calls with the keepChanges parameter set to true. This might be optimised however for the case where the persisted branch has not yet been created (i.e. org.apache.jackrabbit.oak.spi.state.AbstractNodeStor

RE: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Marcel Reutegger
Hi, > Purging (should) only happen for refresh calls with the keepChanges > parameter set to true. This might be optimised however for the case > where the persisted branch has not yet been created (i.e. > org.apache.jackrabbit.oak.spi.state.AbstractNodeStoreBranch.InMemory) by > tweaking the way

Re: Package deployment check for locking leading to large no of small branch commits

2013-12-18 Thread Michael Dürig
On 17.12.13 6:26 , Chetan Mehrotra wrote: 6. Session refresh triggers purge of current pending changes leading to a branch commit Purging (should) only happen for refresh calls with the keepChanges parameter set to true. This might be optimised however for the case where the persisted

RE: Package deployment check for locking leading to large no of small branch commits

2013-12-17 Thread Marcel Reutegger
Hi, > On Tue, Dec 17, 2013 at 12:26 PM, Chetan Mehrotra > wrote: > > So not sure where the fix is to be done. Should lock check be also > > restricted to version at which session is pinned or one needs to be at > > latest revision? > > Since the lock checks are normally used to synchronize work

Re: Package deployment check for locking leading to large no of small branch commits

2013-12-17 Thread Jukka Zitting
Hi, On Tue, Dec 17, 2013 at 12:26 PM, Chetan Mehrotra wrote: > So not sure where the fix is to be done. Should lock check be also > restricted to version at which session is pinned or one needs to be at > latest revision? Since the lock checks are normally used to synchronize work across session

Package deployment check for locking leading to large no of small branch commits

2013-12-17 Thread Chetan Mehrotra
Hi, While profiling slow deployment of some packages on CQ (using Oak) running on Mongo I noticed that its creating lots of small branch commits. Looking into the invocation flow following appears to be happening 1. Package contains lots of small nodes (~7000) 2. DocViewSAXImporter.createNode cre