From 7b4dd972a76ca15dd8c8d9c0eafdedd2ca18b943 Mon Sep 17 00:00:00 2001
From: Vanessasaurus <814322+vsoch@users.noreply.github.com>
Date: Thu, 30 Nov 2023 11:17:52 -0700
Subject: [PATCH] add support for getting spot instance prices (#35)

* add support for getting spot instance prices
* excessive logging
* update test selector to use name, prices likely nto available

Signed-off-by: vsoch <vsoch@users.noreply.github.com>
---
 CHANGELOG.md                       |  1 +
 cloudselect/cloud/aws/client.py    | 78 ++++++++++++++++++++++++++++++
 cloudselect/tests/test_selector.py |  5 +-
 cloudselect/version.py             |  2 +-
 4 files changed, 83 insertions(+), 3 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 3df934e..79b8550 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -14,6 +14,7 @@ and **Merged pull requests**. Critical items to know are:
 The versions coincide with releases on pip. Only major versions will be released as tags on Github.
 
 ## [0.0.x](https://github.com/converged-computing/cloud-select/tree/main) (0.0.x)
+ - support for aws spot prices (function to request) (0.0.23)
  - aws prices no longer works to receive an empty NextToken (0.0.22)
  - fix one-off error for table listing (0.0.21)
    - support for spot-instance example
diff --git a/cloudselect/cloud/aws/client.py b/cloudselect/cloud/aws/client.py
index 8a51d6b..def6f20 100644
--- a/cloudselect/cloud/aws/client.py
+++ b/cloudselect/cloud/aws/client.py
@@ -7,6 +7,7 @@
 import random
 import re
 import time
+from datetime import datetime
 
 import cloudselect.utils as utils
 from cloudselect.logger import logger
@@ -94,6 +95,83 @@ def prices(self):
         print()
         return self.load_prices(prices)
 
+    def spot_prices(self, instances):
+        """
+        Get spot prices for a set of instances and availability zones
+        """
+        from botocore.exceptions import ClientError
+
+        # Filter down to those that support spot
+        instances = [x for x in instances.data if "spot" in x["SupportedUsageClasses"]]
+        names = [x["InstanceType"] for x in instances]
+
+        # Ensure we have services (if cache_only True this won't be set)
+        if not self.has_instance_auth:
+            self._set_services()
+
+        retries = 0
+        prices = {}
+        next_token = ""
+        now = datetime.now()
+        print(f"Getting latest spot prices for {len(names)} instances...")
+        while True:
+            try:
+                response = self.ec2_client.describe_spot_price_history(
+                    InstanceTypes=names,
+                    ProductDescriptions=["Linux/UNIX"],
+                    NextToken=next_token,
+                    MaxResults=1000,
+                    StartTime=now,
+                )
+            except ClientError as err:
+                if "Rate exceeded" not in err.args[0] or retries > self.max_retries:
+                    raise
+                retries += 1
+                sleep = self.min_sleep_time * random.randint(1, 2**retries)
+                time.sleep(sleep)
+                continue
+
+            if not response.get("NextToken"):
+                break
+
+            next_token = response.get("NextToken")
+
+            # Filter down to latest price for each availability zone
+            for price in response["SpotPriceHistory"]:
+                instance_type = price["InstanceType"]
+                zone = price["AvailabilityZone"]
+
+                # Organize by instance type -> availability zone
+                if instance_type not in prices:
+                    prices[instance_type] = {}
+                if zone not in prices[instance_type]:
+                    prices[instance_type][zone] = {}
+
+                # If we already have one, determine if newer
+                if "Timestamp" in prices[instance_type][zone]:
+                    current = prices[instance_type][zone]["Timestamp"]
+                    contender = price["Timestamp"]
+
+                    # Only update if the contender is newer (more recent)
+                    if contender > current:
+                        prices[instance_type][zone] = price
+
+                # We haven't seen this combination yet!
+                else:
+                    prices[instance_type][zone] = price
+
+        # Convert timestamp to strings
+        count = 0
+        for instance_type, spot_prices in prices.items():
+            for zone, spot_price in spot_prices.items():
+                count += 1
+                prices[instance_type][zone]["Timestamp"] = str(
+                    prices[instance_type][zone]["Timestamp"]
+                )
+
+        print(f"Found {count} total aws spot prices")
+        return prices
+
     def instances(self):
         """
         Use the API to retrieve (and return) instances within a set of regions.
diff --git a/cloudselect/tests/test_selector.py b/cloudselect/tests/test_selector.py
index af97d49..cfc8e36 100644
--- a/cloudselect/tests/test_selector.py
+++ b/cloudselect/tests/test_selector.py
@@ -25,9 +25,10 @@ def test_selector(tmp_path, cloud, resources):
     Test our selector (non-interactive)
     """
     selector = selectors.InstanceSelector(cloud=cloud)
-    instances = selector.select_instance(resources, interactive=False)
+
+    # Google prices aren't reliable, so sort by name
+    instances = selector.select_instance(resources, interactive=False, sort_by="name")
     assert len(instances) > 10
-    assert instances[0]["price"] < instances[-1]["price"]
 
     # Ensure each without our range!
     for instance in instances:
diff --git a/cloudselect/version.py b/cloudselect/version.py
index 1a95c72..41a4308 100644
--- a/cloudselect/version.py
+++ b/cloudselect/version.py
@@ -3,7 +3,7 @@
 #
 # SPDX-License-Identifier: (MIT)
 
-__version__ = "0.0.22"
+__version__ = "0.0.23"
 AUTHOR = "Vanessa Sochat"
 EMAIL = "vsoch@users.noreply.github.com"
 NAME = "cloud-select-tool"