pg.ddx.io aws/aws-advanced-jdbc-wrapper GitHub issues and pull requests
help / color / mirror / Atom feed[aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
15+ messages / 2 participants
[nested] [flat]
* [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 21:21 "sophia-bq (@sophia-bq)" <noreply+sophia-bq@github.com>
0 siblings, 0 replies; 15+ messages in thread
From: sophia-bq (@sophia-bq) @ 2026-11-16 21:21 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
### Summary
<!--- General summary / title -->
### Description
<!--- Details of what you changed -->
### Additional Reviewers
<!-- Any additional reviewers -->
### By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
inline patch 1: docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md, wrapper/src/test/java/integration/container/tests/KmsWriteShapesIntegrationTest.java (+51/-4 lines) download inline-1.patch
diff --git a/docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md b/docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md
index 511adfa59..a3c104253 100644
--- a/docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md
+++ b/docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md
@@ -848,13 +848,56 @@ The plugin needs only `kms:Decrypt` on the master keys its columns use:
Grant the actions `KeyManagementUtility` uses (see [AWS Requirements](#aws-requirements)) only to the credentials that set up and rotate the columns.
+### Separating the runtime role from the admin role
+
+Run the application and the setup and rotation tooling under different credentials, so that a compromised application cannot change which keys protect a column.
+
+| | Runtime (application) | Admin (`KeyManagementUtility`, DDL) |
+|---|---|---|
+| AWS | `kms:Decrypt` only, on the specific master key ARNs the columns use | `kms:GenerateDataKey` and `kms:Decrypt` on those ARNs, plus `kms:DescribeKey` for `validateMasterKey()`. `kms:CreateKey` and `kms:CreateAlias` only if it creates master keys |
+| Database | `SELECT` on `key_storage` and `encryption_metadata`; normal DML on the application tables | Write access to `key_storage` and `encryption_metadata`; DDL and trigger creation on the application tables |
+
+The runtime role must not have `kms:GenerateDataKey`, `kms:Encrypt`, `kms:CreateKey` or `"Resource": "*"`, and must not be able to write the metadata tables. Without write access to `key_storage`, a compromised application cannot point a column at a different master key or HMAC key, and the `encryptionAllowedMasterKeyArns` list and the trigger stay meaningful. Use a separate admin credential, held by your deployment pipeline or an operator, and do not ship it with the application.
+
+**PostgreSQL:**, with `app_runtime` as the application's role and `app_admin` as the admin role (replace `encrypt` with your metadata schema):
+
+```sql
+-- Metadata: read-only for the application
+REVOKE ALL ON ALL TABLES IN SCHEMA encrypt FROM PUBLIC;
+REVOKE ALL ON ALL TABLES IN SCHEMA encrypt FROM app_runtime;
+GRANT USAGE ON SCHEMA encrypt TO app_runtime;
+GRANT SELECT ON encrypt.key_storage, encrypt.encryption_metadata TO app_runtime;
+
+-- Metadata: writable only by the admin role
+GRANT USAGE ON SCHEMA encrypt TO app_admin;
+GRANT SELECT, INSERT, UPDATE, DELETE ON encrypt.key_storage, encrypt.encryption_metadata TO app_admin;
+GRANT USAGE, SELECT ON ALL SEQUENCES IN SCHEMA encrypt TO app_admin;
+```
+
+The trigger function runs with the privileges of the user whose statement fires it, so every role that writes an encrypted column, not only `app_runtime`, needs the `SELECT` grants above. Roles that never write encrypted columns need no access to the schema.
+
+**MySQL:**, with `'app_runtime'@'%'` and `'app_admin'@'%'` (replace `encrypt` with your metadata schema):
+
+```sql
+REVOKE ALL PRIVILEGES ON encrypt.* FROM 'app_runtime'@'%';
+GRANT SELECT ON encrypt.key_storage TO 'app_runtime'@'%';
+GRANT SELECT ON encrypt.encryption_metadata TO 'app_runtime'@'%';
+
+GRANT SELECT, INSERT, UPDATE, DELETE ON encrypt.key_storage TO 'app_admin'@'%';
+GRANT SELECT, INSERT, UPDATE, DELETE ON encrypt.encryption_metadata TO 'app_admin'@'%';
+```
+
+`REVOKE` fails on MySQL with error 1141 if the account does not hold the privilege being revoked, so skip it for an account you have not granted anything to. MySQL routines and triggers run with the privileges of their `DEFINER`, which is the account that installed them, the shipped script sets no `SQL SECURITY` clause. That account, the admin role, therefore needs `SELECT` on the metadata tables and the privileges to create routines and triggers (see step 3). Writers of an encrypted column need no privileges on the metadata schema on MySQL; only the plugin's own connection needs the `SELECT` grants above.
+
+Verify the separation by connecting as the runtime role and confirming that `INSERT INTO encrypt.key_storage ...` and `UPDATE encrypt.encryption_metadata ...` fail with a permission error, and that `aws kms generate-data-key` with the runtime credentials is denied.
+
### Data Protection
- Encrypted data is stored as binary data in the database
- The original data never leaves your application - encryption/decryption happens locally using data keys from KMS
- Only encryption keys are managed by AWS KMS, not the actual data
- Use different KMS keys for different environments (dev, staging, prod)
-- Restrict access to the tables in the metadata schema. The HMAC keys in `key_storage` are stored in plaintext so that the triggers can read them. A user that can both read `key_storage` and write an encrypted column can compute a valid HMAC for any value, so the trigger stops plaintext that reaches the column by mistake, not a user who deliberately misuses that access. The application's database user needs `SELECT` on `key_storage` and `encryption_metadata`. The PostgreSQL trigger function runs with the privileges of the user whose statement fires it, so every user that writes an encrypted column needs `SELECT` on both tables as well. Only the user that runs `KeyManagementUtility` needs to write to them
+- Restrict access to the tables in the metadata schema. The HMAC keys in `key_storage` are stored in plaintext so that the triggers can read them. A user that can both read `key_storage` and write an encrypted column can compute a valid HMAC for any value, so the trigger stops plaintext that reaches the column by mistake, not a user who deliberately misuses that access. The application's database user needs `SELECT` on `key_storage` and `encryption_metadata`. On PostgreSQL the trigger function runs with the privileges of the user whose statement fires it, so every user that writes an encrypted column needs `SELECT` on both tables as well. On MySQL the trigger runs as its definer, so writers need no access to the metadata tables. Only the user that runs `KeyManagementUtility` needs to write to them (see [Separating the runtime role from the admin role](#separating-the-runtime-role-from-the-admin-role))
- Always use SSL/TLS for database connections
- Enable CloudTrail logging for KMS operations
@@ -985,6 +1028,7 @@ JOIN encrypt.key_storage ks ON em.key_id = ks.id;
- Create the HMAC validation trigger on every encrypted column (see [step 7 of Setup](#7-add-hmac-validation-triggers))
- Use IAM roles instead of hardcoded credentials or access keys
- Apply least-privilege IAM policies
+ - Run the application under a runtime role with only `kms:Decrypt` and `SELECT` on the metadata schema, and keep a separate admin credential for `KeyManagementUtility` (see [Separating the runtime role from the admin role](#separating-the-runtime-role-from-the-admin-role))
- Restrict access to the metadata schema, and regularly audit the encryption metadata
3. **Performance**
- Use connection pooling
diff --git a/wrapper/src/test/java/integration/container/tests/KmsWriteShapesIntegrationTest.java b/wrapper/src/test/java/integration/container/tests/KmsWriteShapesIntegrationTest.java
index 05a6c6b32..01ed9babb 100644
--- a/wrapper/src/test/java/integration/container/tests/KmsWriteShapesIntegrationTest.java
+++ b/wrapper/src/test/java/integration/container/tests/KmsWriteShapesIntegrationTest.java
@@ -101,10 +101,13 @@ public class KmsWriteShapesIntegrationTest {
private static final String EMAIL = "email";
private static final String UNENCRYPTABLE = "not bind parameters";
/**
- * A plaintext longer than the PostgreSQL encrypted_data domain's minimum length, so that it is the
- * HMAC check that rejects it rather than the domain's length check.
+ * A plaintext at least 61 bytes long, the minimum length of a valid payload (32 HMAC + 1 type + 12 IV
+ * + 16 GCM tag). A shorter one is rejected by the PostgreSQL domain's 45-byte check or by the
+ * trigger's length check ("Invalid encrypted data length") before the HMAC is compared, so it would
+ * not prove that the HMAC check is what rejects it.
*/
- private static final String LONG_PLAINTEXT = "plaintext-that-is-long-enough-to-pass-the-length-check";
+ private static final String LONG_PLAINTEXT =
+ "plaintext-that-is-long-enough-to-pass-both-the-domain-and-trigger-length-checks";
private static final String HMAC_FAILURE = "HMAC verification failed";
private static final SecureRandom RANDOM = new SecureRandom();
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 21:34 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 21:34 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:890)
nit: "Writers of an encrypted column need no privileges on the metadata schema" reads as if it includes the application, but the plugin's connection is itself a writer and always needs `SELECT`. Could we separate the two reasons the tables are read?
On MySQL the trigger reads the metadata as its definer, so a role that writes an encrypted column without going through the plugin (for example, a batch job or the `mysql` client) needs no privileges on the metadata schema. The plugin still reads both tables itself, so the database user it connects as always needs the `SELECT` grants above.
While you're editing this paragraph, there's also a comma splice in "...which is the account that installed them, the shipped script sets no `SQL SECURITY` clause." Parentheses would fix it: "...which is the account that installed them (the shipped script sets no `SQL SECURITY` clause)."
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 21:53 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 21:53 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:858)
The admin column doesn't include `SELECT` and `UPDATE` on the application tables, but `KeyManagementUtility` needs both to re-encrypt.
```suggestion
| Database | `SELECT` on `key_storage` and `encryption_metadata`; normal DML on the application tables | Write access to `key_storage` and `encryption_metadata`; `SELECT` and `UPDATE` on the encrypted columns' tables (re-encryption reads and rewrites every row); DDL and trigger creation on the application tables |
```
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 21:56 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 21:56 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:887)
The MySQL admin block grants only DML on the metadata tables. The paragraph below says the admin also creates the routines and triggers (steps 3 and 7) and runs re-encryption, but the block doesn't show those grants.
```suggestion
GRANT SELECT, INSERT, UPDATE, DELETE ON encrypt.key_storage TO 'app_admin'@'%';
GRANT SELECT, INSERT, UPDATE, DELETE ON encrypt.encryption_metadata TO 'app_admin'@'%';
-- Routines and triggers (steps 3 and 7) and re-encryption, in the database that holds your tables
GRANT CREATE ROUTINE, ALTER ROUTINE, EXECUTE ON app_db.* TO 'app_admin'@'%';
GRANT SELECT, UPDATE, TRIGGER ON app_db.users TO 'app_admin'@'%';
```
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 22:03 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 22:03 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:860)
This says the trigger "stays meaningful" for a compromised application. That only holds if the runtime role can't remove the trigger. The verification step at line 892 doesn't check for this, so a setup that can still be bypassed passes it. Could we add something like:
```suggestion
The runtime role must not have `kms:GenerateDataKey`, `kms:Encrypt`, `kms:CreateKey` or `"Resource": "*"`, and must not be able to write the metadata tables. Without write access to `key_storage`, a compromised application cannot point a column at a different master key or HMAC key, and the `encryptionAllowedMasterKeyArns` list and the trigger stay meaningful. The runtime role must also not own the application tables or hold the `TRIGGER` privilege on them, since that lets it drop or disable the HMAC triggers; create those tables as the admin role, or move ownership with `ALTER TABLE ... OWNER TO app_admin` on PostgreSQL. Use a separate admin credential, held by your deployment pipeline or an operator, and do not ship it with the application.
```
and add "and that `DROP TRIGGER` on an encrypted table fails" to the verification sentence at line 892?
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 22:08 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 22:08 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:867)
These `REVOKE` statements only remove direct grants on the `encrypt` objects. The runtime role can still write the metadata tables in several other ways:
- PostgreSQL: if `app_runtime` owns the `encrypt` tables, it can grant itself the privileges again. Step 4 doesn't say which role creates them, so this is likely when the app user ran step 4. Membership in `rds_superuser`, in `app_admin`, or in `pg_write_all_data` (PG14+) also gives write access that this block doesn't remove.
- MySQL (line 882): `REVOKE ALL PRIVILEGES ON encrypt.*` doesn't touch global grants (`ON *.*`, for example when the app connects as the RDS master user) or table-level grants such as an earlier `GRANT INSERT ON encrypt.key_storage`.
In all of these cases `INSERT INTO encrypt.key_storage` still succeeds after the block runs. Could we add a short note, for example: "Make the admin role the owner of the metadata tables (`ALTER TABLE encrypt.key_storage OWNER TO app_admin;` and the same for `encryption_metadata`), and make sure the runtime role is not a superuser or a member of the admin role. On MySQL, check `SHOW GRANTS FOR 'app_runtime'@'%'` for global and table-level privileges and revoke those too."
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 22:10 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 22:10 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:890)
On MySQL the triggers and routines run as their `DEFINER`, the admin account, so every write to an encrypted table depends on that account staying in place. The previous paragraph tells readers to treat the admin credential as deployment-only, so it's easy to clean it up afterwards. If someone drops `app_admin`, or cuts its grants back to the minimum after setup, every `INSERT` / `UPDATE` on an encrypted table fails.
Could we add a warning, for example: "Do not drop the definer account or revoke its `SELECT` on the metadata tables while the triggers exist. To retire it, recreate the routines and triggers under another definer first. Locking the account (`ACCOUNT LOCK`) only blocks logins and is safe."
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-16 22:11 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-16 22:11 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:877)
The block above revokes from `PUBLIC` but grants `SELECT` back only to `app_runtime`. Any other role that writes an encrypted column and used to read the metadata through `PUBLIC` (an ETL or batch role, for example) then fails inside `validate_encrypted_data_hmac` with `permission denied for table key_storage`.
This paragraph mentions the requirement, but people tend to copy the SQL block as-is. A commented placeholder line in the block would make this harder to miss:
```sql
-- Repeat for every other role that writes an encrypted column
GRANT SELECT ON encrypt.key_storage, encrypt.encryption_metadata TO <other_writer_role>;
```
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-17 20:53 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-17 20:53 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:874)
The schema should be transferred to `app_admin` first, not just the tables. In PostgreSQL a schema's owner can drop any table in it, even one owned by another role. So if `app_runtime` created `encrypt` in step 2, it can still run `DROP TABLE encrypt.key_storage CASCADE` and re-create the metadata with its own HMAC key. The same applies to `users` if `app_runtime` owns its schema. On PostgreSQL 15 and later, `public` is owned by whoever owns the database.
Moving the schema first also fixes the "member of both roles" option below. That role can only run `ALTER TABLE ... OWNER TO app_admin` if `app_admin` has `CREATE` on the schema. Otherwise it fails with `permission denied for schema encrypt`.
```suggestion
ALTER SCHEMA encrypt OWNER TO app_admin;
ALTER TABLE encrypt.key_storage OWNER TO app_admin;
ALTER TABLE encrypt.encryption_metadata OWNER TO app_admin;
-- The admin role owns the application table so that the runtime role cannot drop its triggers.
-- Also make sure app_runtime does not own the table's schema or the database.
ALTER TABLE users OWNER TO app_admin;
```
Could the verification step at line 921 also check that `DROP TABLE encrypt.key_storage` fails?
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-17 20:53 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-17 20:53 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:874)
nit: Once `users` is owned by `app_admin`, migrations that alter it (Flyway, Liquibase, Rails) can no longer run as `app_runtime`. Many apps run migrations as the application user, so it might be worth one sentence saying that migrations touching tables with encrypted columns have to run as the admin role.
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-18 00:22 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-18 00:22 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md)
nit: Now that the block starts with `ALTER SCHEMA`, a role that is neither the master user nor a member of both roles fails on that first statement with `must be owner of schema encrypt`, not `must be owner of table key_storage`. Could we update the quoted error, e.g. "A role that is neither fails with `must be owner of schema encrypt` (or `must be owner of table key_storage`)"?
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-18 00:23 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-18 00:23 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
(on docs/using-the-jdbc-driver/using-plugins/UsingTheKmsEncryptionPlugin.md:873)
nit: "transfer them with ALTER SCHEMA ... OWNER TO app_admin" also covers the database, but a database needs `ALTER DATABASE ... OWNER TO app_admin`.
```suggestion
-- by the database owner) or the database; if it does, transfer them with ALTER SCHEMA ... OWNER TO app_admin
-- and ALTER DATABASE ... OWNER TO app_admin
```
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244 (approved): docs: kms least privilege
@ 2026-11-18 19:05 "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: aaron-congo (@aaron-congo) @ 2026-11-18 19:05 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-18 19:31 "sophia-bq (@sophia-bq)" <noreply+sophia-bq@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: sophia-bq (@sophia-bq) @ 2026-11-18 19:31 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
@sophia-bq closed this.
^ permalink raw [nested|flat] 15+ messages in thread
* Re: [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege
@ 2026-11-18 19:31 "sophia-bq (@sophia-bq)" <noreply+sophia-bq@github.com>
13 siblings, 0 replies; 15+ messages in thread
From: sophia-bq (@sophia-bq) @ 2026-11-18 19:31 UTC (permalink / raw)
To: aws/aws-advanced-jdbc-wrapper <noreply+aws-aws-advanced-jdbc-wrapper@github.com>
@sophia-bq merged this (commit 216490d167f9).
^ permalink raw [nested|flat] 15+ messages in thread
end of thread, other threads:[~2026-11-18 19:31 UTC | newest]
Thread overview: 15+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2026-11-16 21:21 [aws/aws-advanced-jdbc-wrapper] PR #2244: docs: kms least privilege "sophia-bq (@sophia-bq)" <noreply+sophia-bq@github.com>
2026-11-16 21:34 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-16 21:53 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-16 21:56 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-16 22:03 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-16 22:08 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-16 22:10 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-16 22:11 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-17 20:53 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-17 20:53 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-18 00:22 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-18 00:23 ` "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-18 19:05 ` Re: [aws/aws-advanced-jdbc-wrapper] PR #2244 (approved): docs: kms least privilege "aaron-congo (@aaron-congo)" <noreply+aaron-congo@github.com>
2026-11-18 19:31 ` "sophia-bq (@sophia-bq)" <noreply+sophia-bq@github.com>
2026-11-18 19:31 ` "sophia-bq (@sophia-bq)" <noreply+sophia-bq@github.com>
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox