Use expanding bindparam in provider_ids_from_rp_ids in_

This was explored as part of discussion with zzzeek on various
ways to make a statement with a large in_ be more performant.

Batching the queries by chunking rp_ids did not help, but an
expanding bindparam did: it makes the surrounding method about
50% faster [1].

bindparam expanding was added in sqlalchemy 1.2.0 so we bump
the requirements here to the latest 1.2.x bugfix release,
1.2.19.

[1] With the caveat that profiling and benchmarking are very
different things, and in this case it is profiling results
which are being analysed.

Change-Id: Ic4e0cdd87f8f2d76b921059ac4bf16a838913abf

lower-constraints.txt

@@ -76,7 +76,7 @@ rfc3986==0.3.1
 Routes==2.3.1
 six==1.10.0
 smmap2==2.0.3
-SQLAlchemy==1.0.10
+SQLAlchemy==1.2.19
 sqlparse==0.2.4
 statsd==3.2.2
 stestr==1.0.0

placement/objects/research_context.py

@@ -331,10 +331,10 @@ def provider_ids_from_rp_ids(context, rp_ids):
         me_to_root, parent,
         me.c.parent_provider_id == parent.c.id)
     sel = sa.select(cols).select_from(me_to_parent)
-    sel = sel.where(me.c.id.in_(rp_ids))
+    sel = sel.where(me.c.id.in_(sa.bindparam('rps', expanding=True)))
 
     ret = {}
-    for r in context.session.execute(sel):
+    for r in context.session.execute(sel, {'rps': list(rp_ids)}):
         ret[r['id']] = r
     return ret
 

requirements.txt

@@ -3,7 +3,7 @@
 # process, which may cause wedges in the gate later.
 
 pbr!=2.1.0,>=2.0.0 # Apache-2.0
-SQLAlchemy!=1.1.5,!=1.1.6,!=1.1.7,!=1.1.8,>=1.0.10 # MIT
+SQLAlchemy>=1.2.19 # MIT
 keystonemiddleware>=4.18.0 # Apache-2.0
 Routes>=2.3.1 # MIT
 WebOb>=1.8.2 # MIT