storage: TestSplitSnapshotRace_SnapshotWins is disabled

[tbg]

The following test appears to have failed:

#21119:

I160809 16:10:59.496234 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
I160809 16:10:59.496523 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
I160809 16:10:59.499416 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
I160809 16:10:59.499679 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
I160809 16:10:59.500355 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
--- FAIL: TestSplitSnapshotRace_SnapshotWins (16.98s)
    <autogenerated>:12: storage/client_split_test.go:1050, condition failed to evaluate within 15s: expected [0 0 0 2 25 25], got [0 0 0 25 25 25]
=== RUN   TestStoreSplitTimestampCacheReadRace
I160809 16:10:59.511950 storage/engine/rocksdb.go:353  opening in memory rocksdb instance
W160809 16:10:59.514467 gossip/gossip.go:1015  not connected to cluster; use --join to specify a connected node
I160809 16:10:59.524527 storage/store.go:2732  system config not yet available
I160809 16:10:59.527705 storage/replica.go:2660  gossiping cluster id {"00000000-0000-0000-0000-000000000000"} from store 1, range 1
I160809 16:10:59.532568 storage/replica.go:2047  store=1:1 range=1 [/Min-/Max): applying command with forced error: storage/replica.go:1972: no-op on empty Raft entry
E160809 16:10:59.533413 storage/replica.go:2082  store=1:1 range=1 [/Min-/Max): error executing raft command: storage/replica.go:1972: no-op on empty Raft entry
I160809 16:10:59.534131 storage/replica_trigger.go:308  store=1:1 range=1 [/Min-/Max): new range lease replica {1 1 1} 1970-01-01 00:00:00 +0000 UTC 900.000123ms following replica {0 0 0} 1970-01-01 00:00:00 +0000 UTC 0s [physicalTime=1970-01-01 00:00:00.000000123 +0000 UTC]
I160809 16:10:59.534206 storage/replica.go:2682  gossiping sentinel from store 1, range 1
--
I160809 16:11:08.725458 stopper.go:408  quiesceing; tasks left:
18     gossip/infostore.go:290
I160809 16:11:08.725584 stopper.go:408  quiesceing; tasks left:
17     gossip/infostore.go:290
--- PASS: Example_rebalancing (0.95s)
FAIL
FAIL    github.com/cockroachdb/cockroach/storage    86.524s
=== RUN   TestBatchBasics
I160809 16:11:18.274805 storage/engine/rocksdb.go:353  opening in memory rocksdb instance
I160809 16:11:18.277250 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
--- PASS: TestBatchBasics (0.01s)
=== RUN   TestBatchRepr
I160809 16:11:18.292548 storage/engine/rocksdb.go:353  opening in memory rocksdb instance
I160809 16:11:18.294597 storage/engine/rocksdb.go:398  closing in-memory rocksdb instance
--- PASS: TestBatchRepr (0.01s)
=== RUN   TestBatchGet
I160809 16:11:18.302996 storage/engine/rocksdb.go:353  opening in memory rocksdb instance

Please assign, take a look and update the issue accordingly.

[tamird]

Seen again https://circleci.com/gh/tamird/cockroach/523

[tamird]

#8987 is the current incarnation of this.

[rjnn]

This doesn't look like the same issue as #8987. In #8987, SnapshotWins is disabled for being flaky - it's SplitWins which is causing the problem. (SnapshotWins was skipped in #8703).

[tamird]

I believe it is the same issue - this failure comes from the shared setup code used by both tests.

[tamird]

Wait, nope, you're right - this is a failure in the actual test.

[tamird]

This test is disabled as of 63e4425.

[bdarnell]

Assigning myself since I'm looking at this test's sibling in #9204.