Skip to content

Latest commit

 

History

History
144 lines (100 loc) · 6.03 KB

recovery-lab.md

File metadata and controls

144 lines (100 loc) · 6.03 KB
Raw

Recovery Lab

Goal

This laboratory aim is to recover the blockchain state from a snapshot. The student will learn to set up a snapshot and take it as reference to run a blockchain either to restart due to a failure or to start a new node.

Workshop instructions

Edit the config.ini file created in Run a Local Producer Node laboratory by adding the producer_api_plugin plugin. Later create the snapshot and replay nodeos from it.

This laboratory assumes the project is running in your local machine

The student must send:

  1. The nodeos.log file created in the step Finally, run nodeos.

Help material

To help you complete this tutorial, you can take a look at the following resource:

Prerequisite

To successfully complete this laboratory, please, make sure you have the following requirements:

Workshop

1. Copy the Run a Local Producer Node laboratory

Copy the Run a Local Producer Node laboratory you have already completed and add the producer_api_plugin plugin in the config.ini file. Once it is done, run again the nodeos as you did in the producer node laboratory.

The final config.ini file is:

enable-stale-production = true
producer-name = eosio
plugin = eosio::producer_plugin

# New plugin to create snapshots
plugin = eosio::producer_api_plugin

2. Generate a Snapshot

To restore a stoped node or start a new one, first you need a snapshot. To do so, run:

  1. Open your terminal.
  2. Paste next command curl -X POST http://127.0.0.1:8888/v1/producer/create_snapshot into the terminal.

By default, snapshots are written to the data/snapshots directory relative to your nodeos data directory.

The resulting snapshot file would be similar to snapshot-0000006105c35d011d0ee0a17f7d30893ee58c3ba5f4be15fd864b5c83ceabfe.bin. Make sure you remember this path, it is going to be needed in the following step.

3. Replay from a Snapshot

For this laboratory, we are going to replay from an empty folder, follow next steps to successfully replay nodeos from a snapshot.

  1. Run the following command to create the working folder.
mkdir ~/recovery-lab
  1. Copy the snapshot you created in step two to the working directory.
cp <SNAPSHOT_NAME>.bin ~/recovery-lab
  1. Create the following variables in your terminal:
SNAPSHOT=latest.bin
CONFIG_DIR=./config
DATA_DIR=./blockchain
  1. Rename the snapshot file for simplicity:
cd ~/recovery-lab
mv <SNAPSHOT_NAME>.bin latest.bin

In case you set a different name than latest.bin for the snapshot file, make sure to update the value for SNAPSHOT variable to the new name.

  1. Create the snapshot folder.
cd ~/recovery-lab
mkdir -p $DATA_DIR/snapshots
  1. Copy the snapshot file located in your working directory to the snapshots directory.
cd ~/recovery-lab
cp latest.bin $DATA_DIR/snapshots/
  1. Create the config folder.
mkdir ~/recovery-lab/$CONFIG_DIR
  1. Copy the latest version of the config.ini file you modified in the step one and paste it into the config folder.
cd ~/recovery-lab
cp <MODIFIED_CONFIG_INI_FILE> $CONFIG_DIR
  1. Finally, run nodeos.
nodeos \
  --config-dir ./config \
  --data-dir $DATA_DIR/ \
  --blocks-dir $DATA_DIR/blocks \
  --snapshot $DATA_DIR/snapshots/$SNAPSHOT \
  >> nodeos.log 2>&1 &

To check if the replay was successfully, run:

tail -f nodeos.log

You should get an output like:

info  2022-01-06T05:48:52.464 thread-0  http_plugin.cpp:877           operator()           ] start listening for http requests
info  2022-01-06T05:48:52.466 thread-0  http_plugin.cpp:983           add_handler          ] add api url: /v1/node/get_supported_apis
info  2022-01-06T05:48:52.470 thread-0  producer_plugin.cpp:2333      produce_block        ] Produced block 6b74690bf476c0c6... #362 @ 2022-01-06T05:48:52.500 signed by eosio [trxs: 0, lib: 361, confirmed: 0]
info  2022-01-06T05:48:52.907 thread-0  producer_plugin.cpp:2333      produce_block        ] Produced block 5e074e7221533e71... #363 @ 2022-01-06T05:48:53.000 signed by eosio [trxs: 0, lib: 362, confirmed: 0]
info  2022-01-06T05:48:53.302 thread-0  producer_plugin.cpp:2333      produce_block        ] Produced block f7b67b9dfc98422a... #364 @ 2022-01-06T05:48:53.500 signed by eosio [trxs: 0, lib: 363, confirmed: 0]
info  2022-01-06T05:48:53.902 thread-0  producer_plugin.cpp:2333      produce_block        ] Produced block 5d87d585b72718ed... #365 @ 2022-01-06T05:48:54.000 signed by eosio [trxs: 0, lib: 364, confirmed: 0]
info  2022-01-06T05:48:54.402 thread-0  producer_plugin.cpp:2333      produce_block        ] Produced block d2ff4ced2b7c1593... #366 @ 2022-01-06T05:48:54.500 signed by eosio [trxs: 0, lib: 365, confirmed: 0]

The starting producing blockchain number should be different to 0, for this example, the starting block is #362. You can validate it in the line info 2022-01-06T05:48:52.470 thread-0 producer_plugin.cpp:2333 produce_block ] Produced block 6b74690bf476c0c6... #362 @ 2022-01-06T05:48:52.500 signed by eosio [trxs: 0, lib: 361, confirmed: 0].

What's next?