Various updates to CI (#331)

The changes included here, at a 10,000 foot view:

- An overhaul of how tests are ran on Amazon Linux 2 on aarch64 platforms. An adjacent git repo (not shown here) also includes a lot of work to make this happen. Basically, we relied on a "pool" of CI runners to sit and wait idle until a new job came through. Now, CI runners are launched ad-hoc, for as many runners that are needed with or without a matrix. This also allows for the re-run of individual failed tasks if required, whereas before this simply would not work
- Re-introduces sccache back into the runs. Previously, PL/Rust and some required dependencies flat out not work with sccache. However, those issues have been addressed
- Introduces a nightly run setup, which for now does two things: 1.) Tests to make sure that the Dockerfile.try setup works as per the documentation, and 2.) Tests that the Debian packages can be both built and installed. These will be ran nightly against the develop branch, and we can certainly add more to it in the future

NOTE: The nightly CI runs will not happen until these changes eventually land in main on next release.

Author: BradyBonnette

.github/scripts/load_cache.sh

@@ -3,7 +3,6 @@
 # Downloads and extracts cache archive from S3.
 #
 # Expects the following environment variables to already exist:
-#  * ARTIFACT_USER_AWS_PROFILE: the profile to use when issuing AWS CLI commands
 #  * AWS_CACHE_BUCKET: the S3 bucket in which to obtain the archive
 #  * HOME: executing user's home directory
 #
@@ -13,13 +12,16 @@
 # Example usage:
 #  . /path/to/plrust/.github/scripts/load_cache.sh
 #  loadcache "some-cache-key-abc123"
+#
+# Note: This assumes the host in which this script is running on has the ability
+# to read and write from the bucket specified by AWS_CACHE_BUCKET
 
 function loadcache() {
   local cache_key="$1"
 
   echo "Checking to see if cache archive exists: $cache_key"
 
-  if aws s3api head-object --profile $ARTIFACT_USER_AWS_PROFILE --bucket $AWS_CACHE_BUCKET --key $cache_key &> /dev/null; then
+  if aws s3api head-object --bucket $AWS_CACHE_BUCKET --key $cache_key &> /dev/null; then
     echo "Cache archive exists for $cache_key -- downloading and extracting now."
 
     mkdir -p $HOME/artifacts/
@@ -28,7 +30,6 @@ function loadcache() {
     echo "Downloadng archive $cache_key and storing to $archive_path"
 
     aws s3api get-object \
-      --profile $ARTIFACT_USER_AWS_PROFILE \
       --bucket $AWS_CACHE_BUCKET \
       --key $cache_key \
       $archive_path

.github/scripts/save_cache.sh

@@ -3,7 +3,6 @@
 # Creates cache archive and uploads to S3.
 #
 # Expects the following environment variables to already exist:
-#  * ARTIFACT_USER_AWS_PROFILE: the profile to use when issuing AWS CLI commands
 #  * AWS_CACHE_BUCKET: the S3 bucket in which to obtain the archive
 #  * HOME: executing user's home directory
 #
@@ -15,6 +14,9 @@
 #  . /path/to/plrust/.github/scripts/save_cache.sh
 #  my_paths=(/path/one /path/two /path/three)
 #  savecache "some-cache-key-abc123" "${my_paths[@]}"
+#
+# Note: This assumes the host in which this script is running on has the ability
+# to read and write from the bucket specified by AWS_CACHE_BUCKET
 
 function savecache() {
   local cache_key="$1"
@@ -23,7 +25,7 @@ function savecache() {
 
   echo "Checking to see if cache archive exists: $cache_key"
 
-  if aws s3api head-object --profile $ARTIFACT_USER_AWS_PROFILE --bucket $AWS_CACHE_BUCKET --key $cache_key &> /dev/null; then
+  if aws s3api head-object --bucket $AWS_CACHE_BUCKET --key $cache_key &> /dev/null; then
     echo "Cache archive exists for $cache_key -- skipping archive creation."
   else
     echo "Cache archive does not exist for $cache_key -- creating archive now."
@@ -38,7 +40,6 @@ function savecache() {
     echo "Created archive $archive_path -- uploading now"
 
     aws s3api put-object \
-      --profile $ARTIFACT_USER_AWS_PROFILE \
       --bucket $AWS_CACHE_BUCKET \
       --key $cache_key \
       --body $archive_path