https://jira.hyperledger.org/browse/FAB-602



The refactoring allows sharing of transaction execution and validation code across goleveldb and couchdb.
This refactoring also allows plugging-in different schemes for various aspects of transaction execution
such as maintianing isolation during simulation, validation and commit.

Change-Id: I0d3fb1cdff99c3f6f991e7b7469920173f194a87
Signed-off-by: manish <manish.sethi@gmail.com>

https://jira.hyperledger.org/browse/FAB-1363

This changeset moves the ChainID() method from the ChainSupport interface
to the ConsenterSupport one. What necessitates this move is that the
Kafka multichain.Chain object returned/expected by HandleChain(), needs
to be able to allocate the resources necessary to keep up with its
respective chain when the Start() method is invoked by the multichain
manager, as the contract of the Chain interface (defined in
chainsupport.go) dictates.

Keeping up with the respective chain means that the Kafka
multichain.Chain object needs to bring up a producer/consumer that
connects to the corresponding chain partition, so it needs to know the
chain ID. The only object passed to us during the HandleChain call is an
object of type multichain.ConsenterSupport. Therefore, this interface
needs to be extended with the ChainID() method.

I've noted the need for this during my review of this changeset:
https://gerrit.hyperledger.org/r/#/c/2763/ (skip to the comment
timestamped 11-27 18:52, and its follow-up comment)

If you want to see how this will look in action for the Kafka consenter
(coming in a follow-up changeset), please see:
https://github.com/kchristidis/fabric/blob/47752ed61fcab1b26207a9e9075c1c793d723912/orderer/kafka/main.go#L128
https://github.com/kchristidis/fabric/blob/47752ed61fcab1b26207a9e9075c1c793d723912/orderer/kafka/main.go#L143



Change-Id: I80590534097a1f4b0f30267ee541a3d3e7f7e3e1
Signed-off-by: Kostas Christidis <kostas@christidis.io>

https://jira.hyperledger.org/browse/FAB-1362



The list of Kafka brokers used for ordering needs to be shared across the
shims (ordering service nodes). This changeset introduces the key and
the getter of that key's value to the sharedconfig package. It also adds
the necessary definition in the orderer protos.

This value is not yet encoded as a configuration item in the genesis
block that the static bootstrapper produces. This will come at a
subsequent changeset, with the introduction of the provisional
bootstrapper.

Change-Id: I6268e79dc25ca2cee8da248ba44c9a572c04d909
Signed-off-by: Kostas Christidis <kostas@christidis.io>

https://jira.hyperledger.org/browse/FAB-1361



Change-Id: If951bb6d76d96f7f519471b76c73b1ec01b86063
Signed-off-by: Kostas Christidis <kostas@christidis.io>

https://jira.hyperledger.org/browse/FAB-1360



This changeset introduces ChainPartition, a construct that will be used
to identify the Kafka topic/partition that the ordering shims should
interact with when dealing with a particular chain.

Note that it encodes the given chain ID in base-16 as a means of escaping
characters that are not allowed in Kafka topic names. All the
restrictions for topic names are linked to in line 38. Filtering for
complying chain IDs should probably happen during transaction filtering
at ingress. This is marked as a TODO, and a JIRA issue has been created
for it.

This changeset merely introduces the ChainPartition construct and does
not integrate it with the rest of the code yet. This will happen in a
follow-up changeset.

Change-Id: I33d4e80758d7dfd7e7d827caa385d0507a39a40a
Signed-off-by: Kostas Christidis <kostas@christidis.io>

https://jira.hyperledger.org/browse/FAB-1359



As we are slowly moving to a setup where the behavior of the orderer is
controlled by the config options captured in the genesis block and the
orderer YAML file (and their overrides via ENV vars), it's time to drop
the flag support that the Kafka orderer provided.

This changeset then:
1. Removes all flags from the Kafka orderer.
2. Adds a "verbose" option to the YAML file to control logging for the
package that we use to interact with the Kafka cluster (sarama).

Additionally it:
3. Prefixes all test-related variables with "test" so as to make tests
more legible and remove ambiguity.
4. Updates the test config object to match the keys of the actual config
object.
5. Adds a helper test envelope constructor function, and moves some
test-related functions around in an effort to consolidate files.

Change-Id: Id749d0b88f62a4212854e18b8c469d90fe2f6877
Signed-off-by: Kostas Christidis <kostas@christidis.io>

https://jira.hyperledger.org/browse/FAB-1358



There is some mismatch between the batchSize type as expressed in the
localconfig package versus the sharedconfig package and the orderer
configuration proto. This changeset fixes that.

This changeset also defines the precision for the integers in the
localconfig package, for symmetry.

Change-Id: I01d86ff80fff4498f683ad960dd443e189bd1599
Signed-off-by: Kostas Christidis <kostas@christidis.io>

Add a new ledger blockstorage index for History
that will map (blocknum,trannum) to the file storage
location for this block transaction

This index will be used for the API  GetTransactionsForKey()
for (chaincode1,key1).  It will do a key range query on chaincode1~key1
to pick up all chaincode1~key1 records.  Results will indicate
the set of (blocknum,trannum) transactions that updated this key.

Change-Id: I81da09e5526d7e2966634c78a03d34011d514442
Signed-off-by: Mari Wade <mariwade@us.ibm.com>

This implements CSCC to manage multichain from configuration transaction.
It is still missing msp integration as that work hasnt been completed.

Change-Id: I930afb39442270469d82f03f081e6d906e45d1fa
Signed-off-by: Binh Q. Nguyen <binhn@us.ibm.com>

This is an initial check-in that will
1) create the history database in couchDB if it does not exist.
2) add the history commit to the LEDGER Commit if history is enabled
3) stores the history in the history database

The History functionality is not enabled unless both
couchDB and History are enabled in the config.

Note that tests will be added to validate the writes
and the actual data written to the database in future
changes.  The tests to validate history is not possible
until the query APIs are put in place in future changes.

Change-Id: Id207007ab5faae957c1e05234e441566a116ea33
Signed-off-by: Mari Wade <mariwade@us.ibm.com>

This commit introduces:
1) A pull.Mediator, an object that offloads
handling of messages and implementation of algo.PullAdapter from
the rest of the gossip code, so that it can be used for any type of
message we want to sync using the pull mechanism.

2) A certStore object, that holds certificates that are going
   to be replicated using the pull mechanism.
3) An addition of an identity message which is a certificate that's
   replicated among peers.

Change-Id: Ia7682c3c9874ee40fa3588706239f914f1e334fb
Signed-off-by: Yacov Manevich <yacovm@il.ibm.com>

This change-sets removes all dependencies to the crytpo/primitives package
from the msp package. The MSP uses now only the BCCSP for its cryptographic
operations.
This change-sets also removes all dependencies to the
primitives.Hash function. The BCCSP is used to replace thos calls.

Change-Id: Ia1432fac29745bb4f42b9acb8334caf69f115b3d
Signed-off-by: Angelo De Caro <adc@zurich.ibm.com>

per pbft protocol, new-view messages for old views must not be processed.

It is not clear if there was a bug, but with this changeset the
implementation is more streamlined and slightly optimized.

Change-Id: Idbd9e9fdb883516dcf142ef7bdf56794313c8352
Signed-off-by: Marko Vukolic <mvu@zurich.ibm.com>

Change-Id: I064999e6f24eedc6b57a36be67176e0cd79cf3b8
Signed-off-by: Greg Haskins <gregory.haskins@gmail.com>

Despite the verification running and returning positive, the merge seems
to have created an import cycle, this changeset fixes that.

Change-Id: I903766efa524f942d8389bebea27ca0b4e373aa5
Signed-off-by: Jason Yellick <jyellick@us.ibm.com>