From 88452a29a5b0934a190a4d6e6634f2eaf56a8c23 Mon Sep 17 00:00:00 2001
From: Gerrit User 4257 <4257@4a232e18-c5a9-48ee-94c0-e04e7cca6543>
Date: Fri, 3 May 2019 16:32:00 +0000
Subject: [PATCH] Update patch set 5

Patch Set 5:

(3 comments)

Patch-set: 5
---
 6b3f2ff94df327038a6a3323db787c08c40cbfa5 | 61 ++++++++++++++++++++++++
 1 file changed, 61 insertions(+)
 create mode 100644 6b3f2ff94df327038a6a3323db787c08c40cbfa5

diff --git a/6b3f2ff94df327038a6a3323db787c08c40cbfa5 b/6b3f2ff94df327038a6a3323db787c08c40cbfa5
new file mode 100644
index 00000000..57740b1d
--- /dev/null
+++ b/6b3f2ff94df327038a6a3323db787c08c40cbfa5
@@ -0,0 +1,61 @@
+{
+  "comments": [
+    {
+      "key": {
+        "uuid": "dfbec78f_5a9e1308",
+        "filename": "specs/train/atomic-extraroute.rst",
+        "patchSetId": 5
+      },
+      "lineNbr": 30,
+      "author": {
+        "id": 4257
+      },
+      "writtenOn": "2019-05-03T16:32:00Z",
+      "side": 1,
+      "message": "From discussion at the PTG: another problem is that routes are identified only by their data (destination/nexthop/router id); there\u0027s no way to uniquely identify the source of a route. So if two things add the same route then if one is deleted then the other will be broken (or alternatively, we could fail if two Heat stacks try to create the same route, but that\u0027s also bad and doesn\u0027t prevent other tools from doing the same).",
+      "revId": "6b3f2ff94df327038a6a3323db787c08c40cbfa5",
+      "serverId": "4a232e18-c5a9-48ee-94c0-e04e7cca6543",
+      "unresolved": false
+    },
+    {
+      "key": {
+        "uuid": "dfbec78f_7aaf5794",
+        "filename": "specs/train/atomic-extraroute.rst",
+        "patchSetId": 5
+      },
+      "lineNbr": 58,
+      "author": {
+        "id": 4257
+      },
+      "writtenOn": "2019-05-03T16:32:00Z",
+      "side": 1,
+      "message": "Experience with SoftwareDeployments is that this leads to massive confusion where you always have to clarify whether you\u0027re talking about a series of ExtraRoute resources vs. a single ExtraRoutes resource. Something like ExtraRouteList might be better.",
+      "range": {
+        "startLine": 57,
+        "startChar": 56,
+        "endLine": 58,
+        "endChar": 6
+      },
+      "revId": "6b3f2ff94df327038a6a3323db787c08c40cbfa5",
+      "serverId": "4a232e18-c5a9-48ee-94c0-e04e7cca6543",
+      "unresolved": false
+    },
+    {
+      "key": {
+        "uuid": "dfbec78f_fa55871e",
+        "filename": "specs/train/atomic-extraroute.rst",
+        "patchSetId": 5
+      },
+      "lineNbr": 60,
+      "author": {
+        "id": 4257
+      },
+      "writtenOn": "2019-05-03T16:32:00Z",
+      "side": 1,
+      "message": "This is probably a good idea, because it allows users to use e.g. the repeat function or just passing data from parameters to create a variable number of routes.\n\nHowever, it\u0027s very difficult for Heat to store a *list* of UUIDs as the physical ID of a resource. So if we\u0027re assigning a UUID to deal with the idempotency problem identified above, then we can really only do this if we have a single UUID for the group.\n\nIn other words, we can only do it this way if the Neutron API also supports a *group* of extra routes as the resource identified by a UUID.",
+      "revId": "6b3f2ff94df327038a6a3323db787c08c40cbfa5",
+      "serverId": "4a232e18-c5a9-48ee-94c0-e04e7cca6543",
+      "unresolved": false
+    }
+  ]
+}
\ No newline at end of file