> On 2012-05-07 16:24:48, John Sirois wrote:
> >

Thanks John, I'm submitting this.


> On 2012-05-07 16:24:48, John Sirois wrote:
> > src/state/leveldb.cpp, lines 1-3
> > <https://reviews.apache.org/r/4981/diff/4/?file=106631#file106631line1>
> >
> >     What is the algorithm you're using in full?

I've added some more text around the example in state.hpp.


> On 2012-05-07 16:24:48, John Sirois wrote:
> > src/state/leveldb.cpp, line 69
> > <https://reviews.apache.org/r/4981/diff/4/?file=106631#file106631line69>
> >
> >     Since you're actually using leveldb as a proper kv store here, the 
> > auto-compactions should work as intended, but consider the forced full 
> > compaction on initialize anyway.  There is nothing preventing me from 
> > abusing this swap api by passing no UUID and a new lexicographically 
> > greater name on every swap.

Added compaction on initialize.


> On 2012-05-07 16:24:48, John Sirois wrote:
> > src/state/state.hpp, line 80
> > <https://reviews.apache.org/r/4981/diff/4/?file=106632#file106632line80>
> >
> >     There's a trick - so this gives you a compile time type constraint on T 
> > and then this unused pointer is probably optimized away as a noop?

Correct.


> On 2012-05-07 16:24:48, John Sirois wrote:
> > src/state/state.hpp, line 143
> > <https://reviews.apache.org/r/4981/diff/4/?file=106632#file106632line143>
> >
> >     'starting at 0' - stale doc

Fixed.


> On 2012-05-07 16:24:48, John Sirois wrote:
> > src/state/zookeeper.cpp, line 171
> > <https://reviews.apache.org/r/4981/diff/4/?file=106633#file106633line171>
> >
> >     So in the steady state case this is blocking fwict and the return type 
> > is misleading.  Is the idea to zk->get goes to zk->aget behind the scenes 
> > at some later date?

Yes, this is blocked on the existing MasterDetector code.


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/4981/#review7643
-----------------------------------------------------------


On 2012-05-04 00:37:51, Benjamin Hindman wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/4981/
> -----------------------------------------------------------
> 
> (Updated 2012-05-04 00:37:51)
> 
> 
> Review request for mesos, John Sirois and Vinod Kone.
> 
> 
> Summary
> -------
> 
> This is the beginning of a the components for saving "state" information, 
> including what slaves are connected, what frameworks are running, etc. I'll 
> be adding a ZooKeeperState implementation soon, and the master and it's 
> components will use it to save state in a distributed way.
> 
> 
> Diffs
> -----
> 
>   src/Makefile.am cd503a8 
>   src/messages/state.hpp PRE-CREATION 
>   src/messages/state.proto PRE-CREATION 
>   src/state/leveldb.cpp PRE-CREATION 
>   src/state/state.hpp PRE-CREATION 
>   src/state/zookeeper.cpp PRE-CREATION 
>   src/tests/base_zookeeper_test.hpp f35acc0 
>   src/tests/state_tests.cpp PRE-CREATION 
>   src/zookeeper/group.cpp da2e147 
> 
> Diff: https://reviews.apache.org/r/4981/diff
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Benjamin
> 
>

Reply via email to