fix: improve error handling in updater command

[adriantpaez]

We're running the rewards updater as a Kubernetes CronJob. We had the following issue in the last run:

{"level":"info","time":"2025-02-19T17:00:20Z","message":"EstimateGasPriceAndLimitAndSendTx: sending txn (submitRoot) with gasTipCap=1000000 gasFeeCap=1030088 gasLimit=108042"}
{"level":"error","ts":1739984432.8322456,"caller":"updater/updater.go:83","msg":"Failed to submit root","error":"Failed to estimate gas: EstimateGasPriceAndLimitAndSendTx: failed to send txn (submitRoot): 504 Gateway Timeout"}
{"level":"info","ts":1739984432.832575,"caller":"cmd/updater.go:60","msg":"Failed to update","error":"Failed to estimate gas: EstimateGasPriceAndLimitAndSendTx: failed to send txn (submitRoot): 504 Gateway Timeout"}

This is an RPC error, but the CronJob's exit code was 0. Our monitoring system, which is configured to notify if the process does not exist successfully, doesn't receive an alert because of this.

I propose to exit with code 1 in this scenario where the TX wasn’t sent instead of 0.

[seanmcgary]

LGTM