[bug] In some cases minting a token starts return a not complete receipt from chain

[Gelassen]

Steps to reproduce

1. Run existing chain via docker
2. Login existing user
3. Try to add any new offer

Expected behaviour:

New item is successfully added

Observed behaviour

In some cases an error dialog appears. Client logs shows exception on parsing response.

More details

Valid response from chain contains a log object which has valuable data used by mobile client. Not complete response (despite on it is still valid response) doesn't have this log object which leads to the issue. Further research shown reset chain and recreate docker images solves the issue. Because of issue with geth docker volumes chain state no fully preserve its state during container stop and start back - it might be a possible cause of this issue.

Bug is tracked to keep in mind it and has been put in the backlog for the future. I need to hurry up to finish PoC, returning back to VM might be an option to overcome this or similar issues if they become bring more troubles.

[Gelassen]

Attempt to make complete chain preservation:

1. Console shows this three folders are used by ethereum:

  - /tmp/geth/chaindata
  - /tmp/geth/ethash
  - /root/.ethash

,but mapping this one /tmp/geth/chaindata cause chain hangs out - not mining blocks.

2. This outdated answer uses /root/.ethereum, but its mapping doesn't solve the issue. Likely it is not relevant to the modern versions of geth

3. Reference code from one boost projects uses volumes mapping, but it uses it for its own purposes and paths which are quite different from mine.

4. Simply mapping a full image via <host machine path>:/root also doesn't preserve chain state...

[Gelassen]

Faced with this bug this week on the private test chain. Redeploy contracts and running tests against this new contracts have solved the issue.