From 6eebcfc7c4255f4ff73c7b47b7cf209fced46b3e Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 09:28:48 +0200
Subject: [PATCH 01/13] Revert Split Tokenize

This is not included in 2024.4.8 ! So code would not work
---
 distributed/client.py               | 2 +-
 distributed/diagnostics/progress.py | 2 +-
 distributed/protocol/serialize.py   | 2 +-
 distributed/scheduler.py            | 2 +-
 distributed/shuffle/_merge.py       | 2 +-
 distributed/shuffle/_rechunk.py     | 2 +-
 distributed/shuffle/_shuffle.py     | 2 +-
 distributed/tests/test_client.py    | 4 ++--
 8 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/distributed/client.py b/distributed/client.py
index b038c0e46fb..75e8d976eec 100644
--- a/distributed/client.py
+++ b/distributed/client.py
@@ -49,7 +49,7 @@
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
 from dask.optimization import SubgraphCallable
-from dask.tokenize import tokenize
+from dask.base import tokenize
 from dask.typing import Key, NoDefault, no_default
 from dask.utils import (
     apply,
diff --git a/distributed/diagnostics/progress.py b/distributed/diagnostics/progress.py
index 1712cc5df32..3fd7bbbd07f 100644
--- a/distributed/diagnostics/progress.py
+++ b/distributed/diagnostics/progress.py
@@ -9,7 +9,7 @@
 
 from tlz import groupby, valmap
 
-from dask.tokenize import tokenize
+from dask.base import tokenize
 from dask.utils import key_split
 
 from distributed.diagnostics.plugin import SchedulerPlugin
diff --git a/distributed/protocol/serialize.py b/distributed/protocol/serialize.py
index 4ad5c3f98a2..300e27631b2 100644
--- a/distributed/protocol/serialize.py
+++ b/distributed/protocol/serialize.py
@@ -15,7 +15,7 @@
 
 import dask
 from dask.sizeof import sizeof
-from dask.tokenize import normalize_token
+from dask.base import normalize_token
 from dask.utils import typename
 
 from distributed.metrics import context_meter
diff --git a/distributed/scheduler.py b/distributed/scheduler.py
index 64ad6d964f0..9b9828f8045 100644
--- a/distributed/scheduler.py
+++ b/distributed/scheduler.py
@@ -55,7 +55,7 @@
 import dask
 import dask.utils
 from dask.core import get_deps, iskey, validate_key
-from dask.tokenize import TokenizationError, normalize_token, tokenize
+from dask.base import TokenizationError, normalize_token, tokenize
 from dask.typing import Key, no_default
 from dask.utils import (
     _deprecated,
diff --git a/distributed/shuffle/_merge.py b/distributed/shuffle/_merge.py
index 0fb403a3f82..735c8f637fa 100644
--- a/distributed/shuffle/_merge.py
+++ b/distributed/shuffle/_merge.py
@@ -8,7 +8,7 @@
 from dask.base import is_dask_collection
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
-from dask.tokenize import tokenize
+from dask.base import tokenize
 
 from distributed.shuffle._arrow import check_minimal_arrow_version
 from distributed.shuffle._core import ShuffleId, barrier_key, get_worker_plugin
diff --git a/distributed/shuffle/_rechunk.py b/distributed/shuffle/_rechunk.py
index 15e3ea1d78a..ef233e4ed36 100644
--- a/distributed/shuffle/_rechunk.py
+++ b/distributed/shuffle/_rechunk.py
@@ -123,7 +123,7 @@
 import dask.config
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
-from dask.tokenize import tokenize
+from dask.base import tokenize
 from dask.typing import Key
 from dask.utils import parse_bytes
 
diff --git a/distributed/shuffle/_shuffle.py b/distributed/shuffle/_shuffle.py
index 154f25c4bab..cced6a50190 100644
--- a/distributed/shuffle/_shuffle.py
+++ b/distributed/shuffle/_shuffle.py
@@ -23,7 +23,7 @@
 import dask
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
-from dask.tokenize import tokenize
+from dask.base import tokenize
 from dask.typing import Key
 
 from distributed.core import PooledRPCCall
diff --git a/distributed/tests/test_client.py b/distributed/tests/test_client.py
index 0016294e708..fd1d9563e1e 100644
--- a/distributed/tests/test_client.py
+++ b/distributed/tests/test_client.py
@@ -43,7 +43,7 @@
 import dask.bag as db
 from dask import delayed
 from dask.optimization import SubgraphCallable
-from dask.tokenize import tokenize
+from dask.base import tokenize
 from dask.utils import get_default_shuffle_method, parse_timedelta, tmpfile
 
 from distributed import (
@@ -1127,7 +1127,7 @@ async def test_scatter_non_list(c, s, a, b):
 
 @gen_cluster(client=True)
 async def test_scatter_tokenize_local(c, s, a, b):
-    from dask.tokenize import normalize_token
+    from dask.base import normalize_token
 
     class MyObj:
         pass

From 24cc26d74b14bf0438da1fad6d0be7e2faaf6f36 Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 09:38:00 +0200
Subject: [PATCH 02/13] Fix #5918

The method clean_dashboard_address will also return an empty string. In this case the start_address should be used which will account for the "host" parameter used when specifying a Cluster

This will fix the bug a firewall popup will be created even if `Localcluster(host=localhost)` is specified
---
 distributed/node.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/node.py b/distributed/node.py
index d41a3409850..5cb45c95f4e 100644
--- a/distributed/node.py
+++ b/distributed/node.py
@@ -145,7 +145,7 @@ def start_http_server(
         http_addresses = clean_dashboard_address(dashboard_address or default_port)
 
         for http_address in http_addresses:
-            if http_address["address"] is None:
+            if http_address["address"] is None or http_address["address"] == '':
                 address = self._start_address
                 if isinstance(address, (list, tuple)):
                     address = address[0]

From 226b860c9549389d9f98ef82c918bb3a94612bcc Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 09:48:07 +0200
Subject: [PATCH 03/13] Revert "Revert Split Tokenize"

This reverts commit 6eebcfc7c4255f4ff73c7b47b7cf209fced46b3e.
---
 distributed/client.py               | 2 +-
 distributed/diagnostics/progress.py | 2 +-
 distributed/protocol/serialize.py   | 2 +-
 distributed/scheduler.py            | 2 +-
 distributed/shuffle/_merge.py       | 2 +-
 distributed/shuffle/_rechunk.py     | 2 +-
 distributed/shuffle/_shuffle.py     | 2 +-
 distributed/tests/test_client.py    | 4 ++--
 8 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/distributed/client.py b/distributed/client.py
index 75e8d976eec..b038c0e46fb 100644
--- a/distributed/client.py
+++ b/distributed/client.py
@@ -49,7 +49,7 @@
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
 from dask.optimization import SubgraphCallable
-from dask.base import tokenize
+from dask.tokenize import tokenize
 from dask.typing import Key, NoDefault, no_default
 from dask.utils import (
     apply,
diff --git a/distributed/diagnostics/progress.py b/distributed/diagnostics/progress.py
index 3fd7bbbd07f..1712cc5df32 100644
--- a/distributed/diagnostics/progress.py
+++ b/distributed/diagnostics/progress.py
@@ -9,7 +9,7 @@
 
 from tlz import groupby, valmap
 
-from dask.base import tokenize
+from dask.tokenize import tokenize
 from dask.utils import key_split
 
 from distributed.diagnostics.plugin import SchedulerPlugin
diff --git a/distributed/protocol/serialize.py b/distributed/protocol/serialize.py
index 300e27631b2..4ad5c3f98a2 100644
--- a/distributed/protocol/serialize.py
+++ b/distributed/protocol/serialize.py
@@ -15,7 +15,7 @@
 
 import dask
 from dask.sizeof import sizeof
-from dask.base import normalize_token
+from dask.tokenize import normalize_token
 from dask.utils import typename
 
 from distributed.metrics import context_meter
diff --git a/distributed/scheduler.py b/distributed/scheduler.py
index 9b9828f8045..64ad6d964f0 100644
--- a/distributed/scheduler.py
+++ b/distributed/scheduler.py
@@ -55,7 +55,7 @@
 import dask
 import dask.utils
 from dask.core import get_deps, iskey, validate_key
-from dask.base import TokenizationError, normalize_token, tokenize
+from dask.tokenize import TokenizationError, normalize_token, tokenize
 from dask.typing import Key, no_default
 from dask.utils import (
     _deprecated,
diff --git a/distributed/shuffle/_merge.py b/distributed/shuffle/_merge.py
index 735c8f637fa..0fb403a3f82 100644
--- a/distributed/shuffle/_merge.py
+++ b/distributed/shuffle/_merge.py
@@ -8,7 +8,7 @@
 from dask.base import is_dask_collection
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
-from dask.base import tokenize
+from dask.tokenize import tokenize
 
 from distributed.shuffle._arrow import check_minimal_arrow_version
 from distributed.shuffle._core import ShuffleId, barrier_key, get_worker_plugin
diff --git a/distributed/shuffle/_rechunk.py b/distributed/shuffle/_rechunk.py
index ef233e4ed36..15e3ea1d78a 100644
--- a/distributed/shuffle/_rechunk.py
+++ b/distributed/shuffle/_rechunk.py
@@ -123,7 +123,7 @@
 import dask.config
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
-from dask.base import tokenize
+from dask.tokenize import tokenize
 from dask.typing import Key
 from dask.utils import parse_bytes
 
diff --git a/distributed/shuffle/_shuffle.py b/distributed/shuffle/_shuffle.py
index cced6a50190..154f25c4bab 100644
--- a/distributed/shuffle/_shuffle.py
+++ b/distributed/shuffle/_shuffle.py
@@ -23,7 +23,7 @@
 import dask
 from dask.highlevelgraph import HighLevelGraph
 from dask.layers import Layer
-from dask.base import tokenize
+from dask.tokenize import tokenize
 from dask.typing import Key
 
 from distributed.core import PooledRPCCall
diff --git a/distributed/tests/test_client.py b/distributed/tests/test_client.py
index fd1d9563e1e..0016294e708 100644
--- a/distributed/tests/test_client.py
+++ b/distributed/tests/test_client.py
@@ -43,7 +43,7 @@
 import dask.bag as db
 from dask import delayed
 from dask.optimization import SubgraphCallable
-from dask.base import tokenize
+from dask.tokenize import tokenize
 from dask.utils import get_default_shuffle_method, parse_timedelta, tmpfile
 
 from distributed import (
@@ -1127,7 +1127,7 @@ async def test_scatter_non_list(c, s, a, b):
 
 @gen_cluster(client=True)
 async def test_scatter_tokenize_local(c, s, a, b):
-    from dask.base import normalize_token
+    from dask.tokenize import normalize_token
 
     class MyObj:
         pass

From 9ffcb8d47192459021bdf145e9cfa07d307cc5bd Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 10:14:11 +0200
Subject: [PATCH 04/13] Linting / pre-commit hooks (pull_request)

---
 distributed/node.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/node.py b/distributed/node.py
index 5cb45c95f4e..c137f22f5b9 100644
--- a/distributed/node.py
+++ b/distributed/node.py
@@ -145,7 +145,7 @@ def start_http_server(
         http_addresses = clean_dashboard_address(dashboard_address or default_port)
 
         for http_address in http_addresses:
-            if http_address["address"] is None or http_address["address"] == '':
+            if http_address["address"] is None or http_address["address"] == "":
                 address = self._start_address
                 if isinstance(address, (list, tuple)):
                     address = address[0]

From 9cd91a7a46e02bddb9b833e61046c3cf5a032638 Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 10:54:48 +0200
Subject: [PATCH 05/13] Specify defaultcase for dashboard address

Iff the dashboard should listen to all interfaces this should be expressed here since empty strings should respect the given host argument
---
 distributed/node.py | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/distributed/node.py b/distributed/node.py
index c137f22f5b9..08a8dd7cd67 100644
--- a/distributed/node.py
+++ b/distributed/node.py
@@ -143,8 +143,13 @@ def start_http_server(
         self.http_server = HTTPServer(self.http_application, ssl_options=ssl_options)
 
         http_addresses = clean_dashboard_address(dashboard_address or default_port)
-
         for http_address in http_addresses:
+            # Handle default case for dashboard address
+            # In case dashboard_address is given, e.g. ":8787"
+            # the address is empty and it is intended to listen to all interfaces
+            if dashboard_address is not None and http_address["address"] == "":
+                http_address["address"] = "0.0.0.0"
+            
             if http_address["address"] is None or http_address["address"] == "":
                 address = self._start_address
                 if isinstance(address, (list, tuple)):

From 93579153f5c3b331939d96bccc72f5173fe96b4b Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 12:29:58 +0200
Subject: [PATCH 06/13] pr linting

---
 distributed/node.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/node.py b/distributed/node.py
index 08a8dd7cd67..56d0a1882c0 100644
--- a/distributed/node.py
+++ b/distributed/node.py
@@ -149,7 +149,7 @@ def start_http_server(
             # the address is empty and it is intended to listen to all interfaces
             if dashboard_address is not None and http_address["address"] == "":
                 http_address["address"] = "0.0.0.0"
-            
+
             if http_address["address"] is None or http_address["address"] == "":
                 address = self._start_address
                 if isinstance(address, (list, tuple)):

From cf3ac827221c559ded9666148ff1862a11db3177 Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 12:48:38 +0200
Subject: [PATCH 07/13] fix for 9cd91a7a

---
 distributed/node.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/node.py b/distributed/node.py
index 56d0a1882c0..b1036ea4648 100644
--- a/distributed/node.py
+++ b/distributed/node.py
@@ -147,7 +147,7 @@ def start_http_server(
             # Handle default case for dashboard address
             # In case dashboard_address is given, e.g. ":8787"
             # the address is empty and it is intended to listen to all interfaces
-            if dashboard_address is not None and http_address["address"] == "":
+            if dashboard_address is None and http_address["address"] == "":
                 http_address["address"] = "0.0.0.0"
 
             if http_address["address"] is None or http_address["address"] == "":

From 90f30d55bd952c7a6b0b4adf3b73a4ec67dab4de Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 14:46:38 +0200
Subject: [PATCH 08/13] Revert "fix for 9cd91a7a"

This reverts commit cf3ac827221c559ded9666148ff1862a11db3177.
---
 distributed/node.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/node.py b/distributed/node.py
index b1036ea4648..56d0a1882c0 100644
--- a/distributed/node.py
+++ b/distributed/node.py
@@ -147,7 +147,7 @@ def start_http_server(
             # Handle default case for dashboard address
             # In case dashboard_address is given, e.g. ":8787"
             # the address is empty and it is intended to listen to all interfaces
-            if dashboard_address is None and http_address["address"] == "":
+            if dashboard_address is not None and http_address["address"] == "":
                 http_address["address"] = "0.0.0.0"
 
             if http_address["address"] is None or http_address["address"] == "":

From 8a77345ea308b9a07a42c9985acff91a4c07178c Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 14:47:32 +0200
Subject: [PATCH 09/13] Fix expected host binding for dashboard

If None is given to dashboard_address, the binding should now take the given parameter on "host" if set.
---
 distributed/tests/test_scheduler.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/tests/test_scheduler.py b/distributed/tests/test_scheduler.py
index cfb6fdefdb8..8f72267e6bf 100644
--- a/distributed/tests/test_scheduler.py
+++ b/distributed/tests/test_scheduler.py
@@ -1969,7 +1969,7 @@ async def test_scheduler_file():
 @pytest.mark.parametrize(
     "dashboard_address,expect",
     [
-        (None, ("::", "0.0.0.0")),
+        (None, ("::", "0.0.0.0","127.0.0.1")),
         ("127.0.0.1:0", ("127.0.0.1",)),
     ],
 )

From cca2b1be073babbf17ce7961992b04d2cc8b3940 Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Tue, 10 Sep 2024 14:50:26 +0200
Subject: [PATCH 10/13] fix linting

---
 distributed/tests/test_scheduler.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/tests/test_scheduler.py b/distributed/tests/test_scheduler.py
index 8f72267e6bf..dada3e3fc95 100644
--- a/distributed/tests/test_scheduler.py
+++ b/distributed/tests/test_scheduler.py
@@ -1969,7 +1969,7 @@ async def test_scheduler_file():
 @pytest.mark.parametrize(
     "dashboard_address,expect",
     [
-        (None, ("::", "0.0.0.0","127.0.0.1")),
+        (None, ("::", "0.0.0.0", "127.0.0.1")),
         ("127.0.0.1:0", ("127.0.0.1",)),
     ],
 )

From 5db15c579260dff8a27c4510a3e535b867931c0e Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Wed, 11 Sep 2024 09:11:04 +0200
Subject: [PATCH 11/13] Fix test for new behaviour

---
 distributed/tests/test_worker.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/distributed/tests/test_worker.py b/distributed/tests/test_worker.py
index 8bfcd8347b9..0a2fed3219a 100644
--- a/distributed/tests/test_worker.py
+++ b/distributed/tests/test_worker.py
@@ -1115,7 +1115,7 @@ async def test_service_hosts_match_worker(s):
 
     async with Worker(s.address, host="tcp://127.0.0.1") as w:
         sock = first(w.http_server._sockets.values())
-        assert sock.getsockname()[0] in ("::", "0.0.0.0")
+        assert sock.getsockname()[0] in ("::", "127.0.0.1")
 
     # See what happens with e.g. `dask worker --listen-address tcp://:8811`
     async with Worker(s.address, host="") as w:

From 21fdf7863705da09f613991759e69640e9075e00 Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Wed, 11 Sep 2024 09:20:46 +0200
Subject: [PATCH 12/13] Add documentation for LocalCluster

---
 distributed/deploy/local.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/distributed/deploy/local.py b/distributed/deploy/local.py
index 69f5d8af35d..db7e864d73d 100644
--- a/distributed/deploy/local.py
+++ b/distributed/deploy/local.py
@@ -62,6 +62,9 @@ class LocalCluster(SpecCluster):
         'localhost:8787' or '0.0.0.0:8787'.  Defaults to ':8787'.
         Set to ``None`` to disable the dashboard.
         Use ':0' for a random port.
+        When using ':8787' the dashboard will bind to the given interface from parameter "host".
+        If "host" is empty, binding will occur on all interfaces '0.0.0.0'.
+        To avoid firewall issues when deploying locally specify "host" parameter to localhost.
     worker_dashboard_address: str
         Address on which to listen for the Bokeh worker diagnostics server like
         'localhost:8787' or '0.0.0.0:8787'.  Defaults to None which disables the dashboard.

From cb9f62cdc7953e1ed65035522baed79b366dcdbb Mon Sep 17 00:00:00 2001
From: Mario Linker <15095261+maldag@users.noreply.github.com>
Date: Thu, 12 Sep 2024 13:55:04 +0200
Subject: [PATCH 13/13] Update distributed/deploy/local.py

Co-authored-by: Hendrik Makait <hendrik@makait.com>
---
 distributed/deploy/local.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/distributed/deploy/local.py b/distributed/deploy/local.py
index db7e864d73d..b7f736e1ed5 100644
--- a/distributed/deploy/local.py
+++ b/distributed/deploy/local.py
@@ -62,9 +62,9 @@ class LocalCluster(SpecCluster):
         'localhost:8787' or '0.0.0.0:8787'.  Defaults to ':8787'.
         Set to ``None`` to disable the dashboard.
         Use ':0' for a random port.
-        When using ':8787' the dashboard will bind to the given interface from parameter "host".
-        If "host" is empty, binding will occur on all interfaces '0.0.0.0'.
-        To avoid firewall issues when deploying locally specify "host" parameter to localhost.
+        When specifying only a port like ':8787', the dashboard will bind to the given interface from the ``host`` parameter.
+        If ``host`` is empty, binding will occur on all interfaces '0.0.0.0'.
+        To avoid firewall issues when deploying locally, set ``host`` to 'localhost'.
     worker_dashboard_address: str
         Address on which to listen for the Bokeh worker diagnostics server like
         'localhost:8787' or '0.0.0.0:8787'.  Defaults to None which disables the dashboard.