Make transaction info cache size configurable.

[keith-turner]

Fluo caches information about failed transactions. This helps speed up the case where multiple row/cols were modified by a failed transaction. Currently this cache size is hard coded. Also the timeout of the cache is hard coded. It would be nice change these to impl properties. Impl properties are ones that are not part of the public API and are tightly coupled to an internal implementation that could change at any moment.

The cache is created in the TxInfoCache constructor. In this constructor it would be nice to do something like :

TxInfoCache(Environment env) {
   long cahceSize = FluoConfigurationImpl.getTxInfoCacheSize(env.getConfiguration());
   long timeOut = FluoConfigurationImpl.getTxInfoCacheTimeout(env.getConfiguration());
   //construct cache using vals above...not sure if should be long or int.. depends on type that cache takes
}

The methods above would have to be added to FluoConfigurationImpl. These methods should probably follow the pattern of FluoConfigurationImpl.getTxCommitMemory()

Possible property names are fluo.impl.tx.failed.cache.size.mb and fluo.impl.tx.failed.cache.expireTime

[kennethmcfarland]

@keith I would like to take a crack at this one if possible simply due to the classes tight coupling with FluoConfiguration.

[keith-turner]

@kpm1985 that would be great if you could work on it

[kennethmcfarland]

YES!!!

…

On Oct 11, 2017 3:09 PM, "Keith Turner" ***@***.***> wrote: @kpm1985 <https://github.com/kpm1985> that would be great if you could work on it — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#938 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Acg-LBNqslDuLwFOn5RW1voOmoa90H6oks5srTy2gaJpZM4P0JH3> .

[kennethmcfarland]

I am going to begin working on this immediately. Thank you for giving me #938. On Oct 11, 2017 3:10 PM, "Kenneth Mcfarland" <kennethpaulmcfarland@gmail.com> wrote:

…

YES!!! On Oct 11, 2017 3:09 PM, "Keith Turner" ***@***.***> wrote: > @kpm1985 <https://github.com/kpm1985> that would be great if you could > work on it > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#938 (comment)>, or mute > the thread > <https://github.com/notifications/unsubscribe-auth/Acg-LBNqslDuLwFOn5RW1voOmoa90H6oks5srTy2gaJpZM4P0JH3> > . >