This function call was previously used to generate mappings from
potential subjects (or SANs) to certificates within the TLS client
object. However, newer Go versions have deprecated this method, instead
building the mapping automatically based on present certificates at
request time. Because the corresponding client configuration field is
not used in Vault (NameToCertificate), it is safe to remove this call
and leave it nil.
See also: 67d894ee65
See also: https://pkg.go.dev/crypto/tls#Config.BuildNameToCertificate
Signed-off-by: Alexander Scheel <alex.scheel@hashicorp.com>
* check for nil signing key on rotation
* add changelog
* Update nil signing key handling
- bypass setting ExpireAt if signing key is nil in rotate
- return err if singing key is nil in signPayload
* add comment; update error msg on signPayload; refactor UT
In doing some testing I found that the listener clusteraddr isn't really used, or at least isn't as important as the top-level clusteraddr setting. As such, go-sockaddr templating needs to be implemented for the top-level `cluster_addr` setting or it's unusable for HA.
Also fix a nil pointer panic I discovered at the same time.
The auth/token/revoke will not error out if the token does not exists, it
always tries to revoke the token and return success to the client whether
or not the token exists. This makes the behavior of
auth/token/revoke-accessor coherent with this and remove the need to
check whether the token still exists.
* go get vault-plugin-secrets-kv@vault-4290-patch-metadata
* add kv metadata patch command
* add changelog entry
* success tests for kv metadata patch flags
* add more kv metadata patch flags tests
* add kv metadata patch cas warning test
* add kv-v2 key metadata patch API docs
* add kv metadata patch to docs
* prevent unintentional field overwriting in kv metadata put cmd
* like create/update ops, prevent patch to paths ending in /
* fix kv metadata patch cmd in docs
* fix flag defaults for kv metadata put
* go get vault-plugin-secrets-kv@vault-4290-patch-metadata
* fix TestKvMetadataPatchCommand_Flags test
* doc fixes
* go get vault-plugin-secrets-kv@master; go mod tidy
* Attempt to address a data race issue within identity store
* Testcase TestIdentityStore_LocalAliasInvalidations identified a data race issue.
* This reverts the previous attempt to address the issue from #13093
* Fixed null token panic from 'v1/auth/token/' endpoints and returned proper error response
* added changelog entry for PR #13233
* changed error message from 'bad token' to 'null token'
* rebased off of main
* Revert "changed error message from 'bad token' to 'null token'"
This reverts commit 381ed9b32c5ddd5e47adb1643ef7e46fb768bc76.
* changed 'bad token' error message to 'invalid token' after revert
* remove unnecessary vault-data folder
* Support clearing an identity alias' custom_metadata
Previously, an update to an entity alias supported updating the
custom_metadata as long as the update was not empty, which makes it
impossible to clear the metadata values completely.
Fixes:
- empty custom_metadata parameters are honoured on entity alias update
- update related tests
- drop dependency on mapstructure
- reformat with gofumpt
* VAULT-1564 report in-flight requests
* adding a changelog
* Changing some variable names and fixing comments
* minor style change
* adding unauthenticated support for in-flight-req
* adding documentation for the listener.profiling stanza
* adding an atomic counter for the inflight requests
addressing comments
* addressing comments
* logging completed requests
* fixing a test
* providing log_requests_info as a config option to determine at which level requests should be logged
* removing a member and a method from the StatusHeaderResponseWriter struct
* adding api docks
* revert changes in NewHTTPResponseWriter
* Fix logging invalid log_requests_info value
* Addressing comments
* Fixing a test
* use an tomic value for logRequestsInfo, and moving the CreateClientID function to Core
* fixing go.sum
* minor refactoring
* protecting InFlightRequests from data race
* another try on fixing a data race
* another try to fix a data race
* addressing comments
* fixing couple of tests
* changing log_requests_info to log_requests_level
* minor style change
* fixing a test
* removing the lock in InFlightRequests
* use single-argument form for interface assertion
* adding doc for the new configuration paramter
* adding the new doc to the nav data file
* minor fix
* Adding support for SHA3 in the transit backend.
* Adds SHA-3 tests for transit sign/verify path. Adds SHA-3 tests for logical system tools path hash functionality. Updates documentation to include SHA-3 algorithms in system tools path hashing.
* Adds changelog entry.
Co-authored-by: robison jacka <robison@packetized.io>
* See what it looks like to replace "master key" with "root key". There are two places that would require more challenging code changes: the storage path `core/master`, and its contents (the JSON-serialized EncodedKeyringtructure.)
* Restore accidentally deleted line
* Add changelog
* Update root->recovery
* Fix test
Co-authored-by: Nick Cabatoff <ncabatoff@hashicorp.com>
* Unify NewHTTPResponseWriter ant NewStatusHeaderResponseWriter to fix ResponseWriter issues
* adding changelog
* removing unnecessary function from the WrappingResponseWriter interface
* changing logical requests responseWriter type
* reverting change to HTTPResponseWriter
* don't log token error on DR Secondary
* stop gauge collector expiration errors on dr secondary
* don't check dr secondary for token create
* see if CI hits panic
* Revert "don't check dr secondary for token create"
This reverts commit c036a1a544d3a20d29d046f1ee239ab1563ce4d9.
* don't check dr secondary for token create
* Revert "see if CI hits panic"
This reverts commit 1e15aa535cac6e4d1684aaf47c8746c094068eb8.
* remove condition on log
* Move the ctx capture outside the goroutine to avoid a race
* refactor the toggleable wrapper to avoid races
* Move the capture back outside the goroutine
* defer
* Add a periodic test of the autoseal to detect loss of connectivity
* Keep the logic adjacent to autoseal
* imports
* typo, plus unnecessary constant time compare
* changelog
* pr feedback
* More feedback
* Add locking and a unit test
* unnecessary
* Add timeouts to encrypt/decrypt operations, capture activeContext before starting loop
* Add a block scope for the timeout
* copy/paste ftl
* Refactor to use two timeouts, and cleanup the repetitive failure code
* Readd 0ing gauge
* use millis
* Invert the unit test logic
