fix: wrong hana single tenant credential connection loop

[kasimos]

Addresses Ticket #1679

Now auth errors in single tenant hana connections will be logged in the debug log and throw an appropriate error, thereby no longer running into a loop and locking the db user

[johannes-vogel]

Suggested change

	if (err.code === 10) {
	// authentication error
	LOG.debug('connection failed:', err)
	throw new Error(`Pool failed connecting`, { cause: err })
	}
	if (attempt < maxRetries) {
	if (err.code !== 10 && attempt < maxRetries) {

I think the error just needs to bubble up. There's no need to log or wrap it.

[kasimos]

This works except for the initial connection when starting a CAP app. There the proposed change will throw:
"[queue] - Failed to fetch targets for flush Error: Pool resource could not be acquired within 10s"
For requests that are after that hitting the service it will correctly show: "[odata] - 500 - SqlError {message: "authentication failed..."