Fix output and gas extraction in JS tracer for Geth

[goodsoft]

JS tracer of internal transactions had two problems:

1. output field wasn't derived correctly due to typo in field names used.
2. gas and gasUsed didn't match the output from built-in callTracer due to weird calculation method.

Typo is fixed now, and gas/gasUsed fields use a more robust method: gas is taken from the first opcode trace in the scope of a call, while gasUsed is calculated using the remaining gas in the last opcode trace.

Upgrading

Internal transactions previously fetched on Geth-based deployments contain wrong data and should be discarded:

UPDATE transactions SET internal_transactions_indexed_at = NULL;

Checklist for your PR

• I added an entry to CHANGELOG.md with this PR
• If I added new functionality, I added tests covering it.
• If I fixed a bug, I added a regression test to prevent the bug from silently reappearing again.
• I checked whether I should update the docs and did so if necessary

[coveralls]

Pull Request Test Coverage Report for Build 4b999ef0-e0f7-466d-a792-7715b86cc7cc

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.3%) to 82.455%

---

Totals
Change from base Build 7cefe622-f206-463b-ab38-d5da6c0ac6a5:	0.3%
Covered Lines:	4488
Relevant Lines:	5443

---

💛 - Coveralls

[vbaranov]

The number of affected chains:

• Goerli
• Rinkeby
• Aerum
• Callisto

[vbaranov]

Actually, a query to initiate retrace of internal transactions should be corrected to this version:

UPDATE transactions SET internal_transactions_indexed_at = NULL where error is null;

because of "error" constraint violation

UPDATE transactions SET internal_transactions_indexed_at = NULL;
ERROR:  new row for relation "transactions" violates check constraint "error"
DETAIL:  Failing row contains (100047, execution reverted, 300000, 20000000000, 22085, \xb8b59b9b51825280c54c60c1ad5432004d3675b26783eb5bf00d5b4853cb41..., 1, \xcd948855, null, 10, 1124540488372167723265742229195508469373525774133861352876470151..., 3194088922991522124357912292974446783935472125024873699554090371..., 0, 28, 0, 2019-04-25 10:56:15.766357, 2019-04-25 10:56:15.766357, \x3a0ab33f75af0ca7bfe8b44a053044963fc4e8f66cae91c2493c5580ccce16..., 2504081, \x8ee7d8bf9b722dfd003146a242bd975a1c524d1c, \xd813419749b3c2cdc94a2f9cfcf154113264a9d6, null, null, null, null).

And also we need to increase timeout for geth indexers, because after implementing this update to half of the geth-based instances of Blockscout, fetcher of internal transactions cannot process 10 internal transactions because of 1-minute timeout. I will add an increase of timeout in a separate PR.

[goodsoft]

Rather, it should set error = NULL as well, because otherwise transactions with non-null error won't be refetched at all, but their gas numbers might be as wrong as the rest.

[vbaranov]

I was actually unsure, that the error status will be refetched if I clear it. Since it will be updated finally, thus, I will update transactions table with:

UPDATE transactions SET internal_transactions_indexed_at = NULL, error = NULL;

Thank you.