byfn test failing intermittently in CI due to peer not
being available when the join channel command is
executed. Need to wait for peer to finsih establishing
connection with couchdb before joining channel.
The exit code was not being correctly captured from the
peer join command and thus the retry logic was not being
executed.
Change-Id: I2c27d17cd769c6b8de1bdcfed263d9b0f758b432
Signed-off-by: Saad Karim <skarim@us.ibm.com>
Docker compose makes volume management easy and transparent to the user.
By defaulting the ledgers to be docker compose volumes, we transparently
get ledger persistence for the ugprade case, and may trivially tear the
volumes down in the docker-compose down command with the addition of the
--volumes flag. This would allow the upgrade scenario to work without
the use of the '-p' flag, and is likely to be generally much cleaner.
Change-Id: I80b36a63370065a7aad0f9c284ec578533123d2b
Signed-off-by: Jason Yellick <jyellick@us.ibm.com>
replace DEBUG level logging w/INFO, but add in
shell debug statements for all fabric commands
for improved UX
Change-Id: I97a19db0373ecaf7a504cb95d4815a11cd64dfed
Signed-off-by: Christopher Ferris <chrisfer@us.ibm.com>
Adding upgrade functionality to byfn.sh to upgrade the fabric network
from v1.0.x to v1.1 and scripts for cli to enable
/Channel, /Channel/Orderer, /Channel/Application capabilities
Usage:
git fetch origin
git checkout origin/release
./byfn.sh -m up -i 1.0.6
git checkout origin/master
./byfn.sh upgrade
Change-Id: I6f53a6db39501e2653dc4c325d3d42d78f463e87
Signed-off-by: Surya <suryalnvs@gmail.com>
Signed-off-by: Jason Yellick <jyellick@us.ibm.com>
This CR adds a new docker-compose-persist.yaml file which creates a
volume mount in the current directory at ./ledgers/<container> and binds
it to the ledger directory for the container.
It also adds a '-p' flag to the byfn.sh script to allow users to include
this new compose file when bringing up the network.
Change-Id: I1ee5faa33c7f0ce18fe7711f5f752d7bf18f117c
Signed-off-by: Jason Yellick <jyellick@us.ibm.com>
Signed-off-by: Surya <suryalnvs@gmail.com>
The first-network tutorial turns on the v1.1 fabric capabilities, and
depends on very recent versions of cryptogen and configtxlator. Users
appear to be running the latest version of first-network with older
binaries, and are experiencing cryptic failures.
This CR adds basic version checking, both for the local binary versions
as well as the binary versions contained within the docker images. If
these versions are mismatched, it prints a warning. If either of the
versions is in a black-list, the byfn.sh will error out.
Change-Id: I1a43ec396c8d1f5a438472cb422d5a7e52e2ef63
Signed-off-by: Jason Yellick <jyellick@us.ibm.com>
Adding IMAGE_TAG option to byfn script to launch specific
version of fabric network
Usage: ./byfn.sh up -i 1.1.0-alpha
Change-Id: I67a1d01257abd7021ea344c354c6e1c97eebf4f7
Signed-off-by: Surya <suryalnvs@gmail.com>
Signed-off-by: Jason Yellick <jyellick@us.ibm.com>
This change brings a new set of scripts and configuration files to the
first-network sample to make it easier for people to follow the new
tutorial on how to add a third org to the network setup in BYFN.
To function properly the new Extend You First Network script (eyfn.sh)
must be run after byfn.sh is run and with the same parameters. So,
valid uses include:
./byfn.sh up
./eyfn.sh up
or
./byfn.sh up -c testchannel -s couchdb -l node
./eyfn.sh up -c testchannel -s couchdb -l node
A single './eyfn.sh down' command is however necessary to take the
whole network down.
Patch-set #2: fixes ./eyfn.sh down
Patch-set #3: removed unused option from Usage and spurious whitespaces
Patch-set #4: added missing test file
Change-Id: I9c926b52f2243dda1c5f9368112c314a6c5c6929
Signed-off-by: Arnaud J Le Hors <lehors@us.ibm.com>
This change fixes a set of related issues.
It fixes the use of the cli timeout setting. It is now merely about
the maximum time spent waiting for a response.
The e2e script is no longer run from docker-compose but by calling
docker exec from byfn.sh and docker-compose now merely has the cli
waiting/sleeping for 1000 seconds.
This not only allows starting the network without automatically
running the e2e script, it also allows byfn.sh to properly end when
the e2e script ends rather than hang. It also makes the tutorial
simpler.
The time the cli container waits/sleeps cannot be changed other than
by editing the config file but this shouldn't be a problem.
This change also fixes the log output of the e2e script so that peers
are consistantly being referenced as in peerX.orgY.
In addition, this changes the usage of byfn.sh so that the -m
parameter is no longer needed. One merely needs to type "./byfn.sh up"
rather than "./byfn.sh -m up". The old way is still supported though,
for human backward compatibility. ;-) And by default the answer to the
prompt for continuation is "y" so the user can simply continue with a
hit of the return key.
Change-Id: I3e2409d81018f97ec9306ce6e1ae916d8fe18fec
Signed-off-by: Arnaud J Le Hors <lehors@us.ibm.com>
The CLI timeout for byfn is currently set to
10000 seconds. This is not a huge issue, but
does mean that the CLI will hang around for
10000 seconds after the e2e is complete.
This simply changes the default timeout to
10 seconds.
Change-Id: I4e163ed248d9e937354e83c0d0143dc6748b6a08
Signed-off-by: Gari Singh <gari.r.singh@gmail.com>
This CR adds two new yaml files that are specific to a new
organization - Org3 - that will join the existing application
channel. Also adds a separate docker-compose that is specific
to the new org - Org3.
Update byfn.sh to remove the Org3 artifacts
Reliant on CR - 15323
Change-Id: I22a08be6f8472f981c4231491b7cae56906b71dd
Signed-off-by: Nick Gaski <ngaski@us.ibm.com>
It's possible to run byfn -m generate multiple
times depending on how you go through the
tutorials, start/stop the network, etc.
The crypto-config folder is deleted when
running byfn -m down and that will stay,
but now each time generateCerts is run the
cyrpto-config folder will be deleted as well.
Change-Id: Ie3851debd76ff0c50c84e14fbca56de2a28bd825
Signed-off-by: Gari Singh <gari.r.singh@gmail.com>
Fabric 1.1 supports javascript chaincode. This changeset
addresses porting of the golang chaincode to node.js
chiancode and the corresponding README files
Change-Id: Iae24e713f16ab3508fe0cc18ee062ffa412b8ba6
Signed-off-by: ratnakar <asara.ratnakar@gmail.com>
Added quotes to $CURRENT_DIR in byfn.sh so that the current directory
can contain spaces and the script will still run properly.
Change-Id: I3853e3398c29c55c46603477fada5db023808431
Issue-id: FAB-5618
Signed-off-by: Ethan Coeytaux <eacoeytaux@gmail.com>
[FAB-5992] Fix error at line 42 of the script.sh in scripts dir,
and add tips in byfn.sh.
Change-Id: Id05277074dc240c06e1b47ac4eedf349e9555cfa
Signed-off-by: Zhangjiong Xuan <xuanzhangjiong@hyperchain.cn>
Remove docker-compose script to fix the byfn e2e test failures
Change-Id: Iee560b73f5fe535120ee888a816646d3aba224ed
Signed-off-by: rameshthoomu <rameshbabu.thoomu@gmail.com>
fixed the unrecognized option on byfn.sh script when
providing the docker-compose file to start the network
Fix Issue # FAB-5603
Change-Id: I74234c53d8f82cbf2678f80aa3a898c3ec51c422
Signed-off-by: dsanchezseco <d.seco@protonmail.com>
Using the -f flag is possible to specify the docker-compose file to use on the
byfn.sh script to have a standar way to test the different configurations for
the samples project. The defautl is 'docker-compose-cli.yaml'
-h output updated as well.
Fix Issue # FAB-5576.
Change-Id: I8766ab930f05d9c4934b149872b9de3a299ff345
Signed-off-by: dsanchezseco <d.seco@protonmail.com>
basic-network:
- moved "crypto-config" to top-level and removed
"network" folder
- added generate.sh to re-gen crypto materials
- docker-compose.yaml to ref v1.0.0 images
- add CLI to docker-compose
- start.sh to only start CA, orderer, peer and couch
- startFabric.sh to run start.sh, then launch CLI
for create channel, install, instantiate, invoke
fabcar:
- moved chaincode to central chaincode folder
- moved "creds" to top-level and removed "network"
folder
- changed to use the basic-network as the target
and removed docker-compose.yaml and crypto-config
- updated package.json to require v1.0.0 modules
- added missing license headers
- restructured to use a central chaincode subdirectory
Change-Id: Ic784d1cf55ea51da5155624f3c38275883de1dca
Signed-off-by: Christopher Ferris <chrisfer@us.ibm.com>
Signed-off-by: Nick Gaski <ngaski@us.ibm.com>
Signed-off-by: Jim Zhang <jzhang@us.ibm.com>
The descriptions for "down" and "restart" are the same as "up".
This patch set changes the descriptions.
Change-Id: I0ee03e5d81faa4ad6a3c43354e8a44648288df90
Signed-off-by: Nao Nishijima <nao.nishijima@hal.hitachi.com>
If you run ./byfn -m up prior to running
./byfn -m generate, things will fail - which
would normally be OK except in the case where
you are running the Docker daemon as root -
which is the default on Linux systems. The issue
is that the Docker daemon will automatically
generate host paths for volumes if they don't exist
and they will be owned by whoever is running the
Docker process resulting in the error reported
in FAB-5038.
To fix this, simply check for the existence of
the crypto-config folder and if it does not
exist run the equivalent commands as the
generate option
Change-Id: Ic5b48682a527951ac6eab5b5cf2d0fc6dbdaad56
Signed-off-by: Gari Singh <gari.r.singh@gmail.com>
and fix incorrect function name in byfn.sh
add *.tx and *.block files to .gitignore
add FABRIC_CFG_PATH setting
Change-Id: Id3457455354b1f2aea1a9661aa89dc677e05124b
Signed-off-by: Christopher Ferris <chrisfer@us.ibm.com>
Signed-off-by: Rameshbabu <rameshbabu.thoomu@gmail.com>
