BFT Block Puller: updatable connection source #4571
Conversation
tock-ibm
force-pushed
the
bft-bp-block-conn-source
branch
from
76b559b
to
9c73888
Compare
tock-ibm
force-pushed
the
bft-bp-block-conn-source
branch
10 times, most recently
from
754fb6c
to
d0f795d
Compare
tock-ibm
force-pushed
the
bft-bp-block-conn-source
branch
4 times, most recently
from
6612036
to
4682dae
Compare
tock-ibm
force-pushed
the
bft-bp-block-conn-source
branch
from
4682dae
to
65539cd
Compare
Signed-off-by: Yoav Tock <tock@il.ibm.com> Change-Id: Id0b80df892595cf2b9d7b7b50bfa9069f0eb6905
tock-ibm
force-pushed
the
bft-bp-block-conn-source
branch
from
65539cd
to
26e1746
Compare
| @@ -0,0 +1,55 @@ | |||
| /* | |||
There was a problem hiding this comment.
don't we have any util package in our entire codebase we can move this function to?
There was a problem hiding this comment.
It got promoted from orderer/common/cluster/util.go to common/deliverclient/util.go, where I plan to put all the stuff that is common to both orderer and peer in the context of the block puller. Can't put it in package common/util nor protoutil because of import cycles. I am open to suggestions.
There was a problem hiding this comment.
I suppose the import cycle is because of github.com/hyperledger/fabric/common/configtx? Then can't you move it there?
There was a problem hiding this comment.
However if you think there are additional things to be placed here, I guess we can leave it as is.
There was a problem hiding this comment.
In terms of import cycles it can be placed in common/configtx, but I think it is not a good place to put code that parses blocks, as that package deals only with individual txs. I prefer to leave it where it is for the time being. I think that during the refactoring that is about to happen when we overhaul the orderer, several additional things will get moved into that util.go file.
Type of change
Description
Make the connection source update when a config block is processed by the block puller, not when it is committed.
Related issues
#4347