feat(sis): Adding deployment_id and gpu_spec_id and index for them to instances table

[dmikhaylovnv]

TL;DR

Adding deployment_id and gpu_spec_id and their indexes for them to instances table

Issues

Resolves NVIDIA/nvcf#11152

Summary by CodeRabbit

• New Features
  • Instances can now be associated with a deployment and GPU specification.
  • Added database indexes to improve lookup performance for these associations.

[coderabbitai]

📝 Walkthrough

Walkthrough

The Cassandra schema adds nullable deployment and GPU specification UUID columns to sis_api.instances. It also adds idempotent custom indexes for both columns.

Changes

Instance lookup schema

Layer / File(s)	Summary
Instance columns and indexes migrations/cassandra/keyspaces/sis_api/07_add_deployment_and_gpu_spec_id.up.sql, migrations/cassandra/keyspaces/sis_api/08_add_indexes_for_deployment_and_gpu_spec_id.up.sql	Adds nullable UUID columns to sis_api.instances and creates custom indexes for deployment and GPU specification lookups.

Estimated code review effort: 1 (Trivial) | ~2 minutes

Suggested reviewers: sanjay-saxena

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title uses the required feat(sis): format and accurately describes the added instance columns and indexes.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/sis_add_columns

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In
`@migrations/cassandra/keyspaces/sis_api/07_add_deployment_and_gpu_spec_id.up.sql`:
- Line 2: Extend migrations/cassandra/tests/test-execute-sqls.sh to cover
migrations 07 and 08: verify both UUID columns, both indexes, migration
ordering, and successful repeat application; the migration SQL files require no
direct changes. If coverage is intentionally omitted, document the reason in the
Pull Request description.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 82eab019-e4ef-4424-b25f-8a2d7e6adcc1

📥 Commits

Reviewing files that changed from the base of the PR and between 399964b and e394713.

📒 Files selected for processing (2)

• migrations/cassandra/keyspaces/sis_api/07_add_deployment_and_gpu_spec_id.up.sql
• migrations/cassandra/keyspaces/sis_api/08_add_indexes_for_deployment_and_gpu_spec_id.up.sql

[coderabbitai]

📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
set -eu

printf '%s\n' '--- migration files ---'
git ls-files 'migrations/cassandra/keyspaces/sis_api/*' | sort

printf '%s\n' '--- repository test and migration references ---'
rg -n -i --glob '!vendor' --glob '!node_modules' \
  'cassandra|migration|migrations|IF NOT EXISTS|StorageAttachedIndex' \
  .github Makefile* README* go.mod go.sum package.json pyproject.toml 2>/dev/null | head -n 250

printf '%s\n' '--- likely test files ---'
git ls-files | rg -i '(^|/)(test|tests|spec|specs)(/|$)|(_test|\.spec|\.test)\.' | head -n 250

Repository: NVIDIA/nvcf

Length of output: 17899

---

🏁 Script executed:

#!/bin/bash
set -eu

printf '%s\n' '--- migration test script ---'
wc -l migrations/cassandra/tests/test-execute-sqls.sh
cat -n migrations/cassandra/tests/test-execute-sqls.sh

printf '%s\n' '--- Cassandra migration tree ---'
git ls-files 'migrations/cassandra/**' | sort

printf '%s\n' '--- relevant migration contents ---'
for f in \
  migrations/cassandra/keyspaces/sis_api/01_init_keyspace.up.sql \
  migrations/cassandra/keyspaces/sis_api/03_init_tables.up.sql \
  migrations/cassandra/keyspaces/sis_api/07_add_deployment_and_gpu_spec_id.up.sql \
  migrations/cassandra/keyspaces/sis_api/08_add_indexes_for_deployment_and_gpu_spec_id.up.sql
do
  echo "### $f"
  cat -n "$f"
done

printf '%s\n' '--- tracked change summary ---'
git diff --stat
git status --short

Repository: NVIDIA/nvcf

Length of output: 31709

---

🏁 Script executed:

#!/bin/bash
set -eu

printf '%s\n' '--- migration documentation ---'
for f in migrations/cassandra/README.md migrations/cassandra/keyspaces/README.md; do
  echo "### $f"
  cat -n "$f"
done

printf '%s\n' '--- migration test references ---'
rg -n -C 4 \
  'test-execute-sqls|migrations/cassandra/tests|execute_sqls\.sh|migrations/cassandra' \
  .github Makefile* BUILD* WORKSPACE* WORKSPACE.bazel 2>/dev/null | head -n 300

printf '%s\n' '--- migration executor ---'
wc -l migrations/cassandra/execute_sqls.sh
cat -n migrations/cassandra/execute_sqls.sh

printf '%s\n' '--- migration ordering and idempotency markers ---'
python3 - <<'PY'
from pathlib import Path

root = Path("migrations/cassandra/keyspaces")
for keyspace in sorted(p.parent.name for p in root.glob("*/01_init_keyspace.up.sql")):
    files = sorted(p.name for p in (root / keyspace).glob("*.up.sql"))
    print(keyspace, ":", ", ".join(files))
    for name in files:
        text = (root / keyspace / name).read_text()
        print(f"  {name}: IF NOT EXISTS={text.upper().count('IF NOT EXISTS')}")
PY

Repository: NVIDIA/nvcf

Length of output: 22912

---

Add coverage for the new Cassandra migrations.

Extend migrations/cassandra/tests/test-execute-sqls.sh to check both UUID columns, both indexes, migration ordering, and repeat application. If tests are intentionally omitted, document the reason in the Pull Request description.

📍 Affects 2 files

• migrations/cassandra/keyspaces/sis_api/07_add_deployment_and_gpu_spec_id.up.sql#L2-L2 (this comment)
• migrations/cassandra/keyspaces/sis_api/08_add_indexes_for_deployment_and_gpu_spec_id.up.sql#L2-L3

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In
`@migrations/cassandra/keyspaces/sis_api/07_add_deployment_and_gpu_spec_id.up.sql`
at line 2, Extend migrations/cassandra/tests/test-execute-sqls.sh to cover
migrations 07 and 08: verify both UUID columns, both indexes, migration
ordering, and successful repeat application; the migration SQL files require no
direct changes. If coverage is intentionally omitted, document the reason in the
Pull Request description.

Source: Coding guidelines

[sanjay-saxena]

@dmikhaylovnv -- Is there a reason why we need separate sql files for these DDLs that are related to the same columns?

[dmikhaylovnv]

@dmikhaylovnv -- Is there a reason why we need separate sql files for these DDLs that are related to the same columns?

I'm following the same approach we have in helenus. My gut feeling is the order of commands in a single file is not guaranteed in some cases while files order is always guaranteed. We need to make sure the columns are created before we add indexes.