Fix tail->repl_offset update in feedReplicationBuffer (#11905)

In #11666, we added a while loop and will split a big reply
node to multiple nodes. The update of tail->repl_offset may
be wrong. Like before #11666, we would have created at most
one new reply node, and now we will create multiple nodes if
it is a big reply node.

Now we are creating more than one node, and the tail->repl_offset
of all the nodes except the last one are incorrect. Because we
update master_repl_offset at the beginning, and then use it to
update the tail->repl_offset. This would have lead to an assertion
during PSYNC, a test was added to validate that case.

Besides that, the calculation of size was adjusted to fix
tests that failed due to a combination of a very low backlog size,
and some thresholds of that get violated because of the relatively
high overhead of replBufBlock. So now if the backlog size / 16 is too
small, we'll take PROTO_REPLY_CHUNK_BYTES instead.

Co-authored-by: Oran Agra <oran@redislabs.com>

1 files changed, 11 insertions, 1 deletions

diff --git a/tests/integration/replication-buffer.tcl b/tests/integration/replication-buffer.tcl
index fe85632ae..2e402480d 100644
--- a/tests/integration/replication-buffer.tcl
+++ b/tests/integration/replication-buffer.tcl
@@ -267,7 +267,7 @@ test {Replica client-output-buffer size is limited to backlog_limit/16 when no r
             wait_for_ofs_sync $master $replica
 
             # Write another key to force the test to wait for another event loop iteration
-            # to give the serverCron a chance to disconnect replicas with COB size exeeeding the limits
+            # to give the serverCron a chance to disconnect replicas with COB size exceeding the limits
             $master set key1 "1"
             wait_for_ofs_sync $master $replica
 
@@ -280,6 +280,16 @@ test {Replica client-output-buffer size is limited to backlog_limit/16 when no r
             }
 
             assert {[status $master sync_partial_ok] == 0}
+
+            # Before this fix (#11905), the test would trigger an assertion in 'o->used >= c->ref_block_pos'
+            test {The update of replBufBlock's repl_offset is ok - Regression test for #11666} {
+                set rd [redis_deferring_client]
+                set replid [status $master master_replid]
+                set offset [status $master repl_backlog_first_byte_offset]
+                $rd psync $replid $offset
+                assert_equal {PONG} [$master ping] ;# Make sure the master doesn't crash.
+                $rd close
+            }
         }
     }
 }