DarrylWong

I'll leave the discussion on if we should add this up to the others since I have no strong opinion/limited insight. But if we do go ahead with this change...

Ran [all the tests](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyGceBazel/14087100?hideProblemsFromDependencies=false&hideTestsFromDependencies=false&expandBuildProblemsSection=true&expandBuildTestsSection=true) that use `internalPgUrl` or `internalAddr` + a [0.2 test run here](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyGceBazel/14086944?hideProblemsFromDependencies=false&hideTestsFromDependencies=false&expandBuildChangesSection=true&expandBuildTestsSection=true) 😁 Gladly accepting alternative name ideas for the constant: https://github.com/cockroachdb/cockroach/blob/5df2a4bbbb12d2849a116067eea8e5a04d7e9424/pkg/roachprod/install/cluster_synced.go#L1611 Most of the diff is...

Ahh, upon further testing, it appears that it's not always okay to default to `localhost` for `InternalPgUrl`. Specifically, when we have a workload only node, it doesn't recognize `localhost` and...

[Reran the affected tests](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyGceBazel/14139919?hideProblemsFromDependencies=false&hideTestsFromDependencies=false&expandBuildChangesSection=true&expandBuildProblemsSection=true&expandBuildTestsSection=true). TFTR! bors r=renatolabs

[Ran the GCE nightly suite here](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyGceBazel/14125018?buildTab=overview&hideProblemsFromDependencies=false&expandBuildTestsSection=true&hideTestsFromDependencies=false&expandBuildChangesSection=true). Only a few related test failures that have been fixed. Rerunning those + preempted tests [here](https://teamcity.cockroachdb.com/buildConfiguration/Cockroach_Nightlies_RoachtestNightlyGceBazel/14160037?buildTab=tests&status=passed) `tpchvec/direct_scans/mt-shared-process` failure was due to a rebase mistake,...

TFTRs! bors r=srosenberg, renatolabs

TFTR! bors r=srosenberg

🥺 bors retry

TFTR!

Looks like this is failing every time, but is usually skipped because Azure doesn't have enough capacity. Seeing this quite often for `westus2`. ``` compute.VirtualMachinesClient#CreateOrUpdate: Failure sending request: StatusCode=0 --...