* fix(proxy): stamp the client alias on a copy of each streamed chunk so pricing sees the deployment model
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(spend): streamed alias matching a capability rule bills the deployment price
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(spend): assert every streamed chunk carries the client alias
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(logging): log the client alias on the priced streamed response, the same as non-streamed
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
(cherry picked from commit 7b432d78d2)
schema.prisma still declares the LiteLLM_SpendLogs (api_key, startTime) and (litellm_call_id) indexes while their migrations are now SELECT 1, so the v1 resolver's drift script recreated both with a blocking CREATE INDEX at boot. The drift script now drops those two statements before it is applied, matching filter_request_log_index_diff on main. The key metadata recovery fixture builds its index directly and the schema sync test exempts the two indexes.
ENFORCE_PRISMA_MIGRATION_CHECK now defaults to true in both the proxy CLI and the standalone migration entrypoint, with ENFORCE_PRISMA_MIGRATION_CHECK=false or --no-enforce_prisma_migration_check as the opt-out. The two LiteLLM_SpendLogs index migrations become SELECT 1 so operators build those indexes online by hand.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): restore pre-config-wins handling of pass-through endpoints
Config-wins (#41779) made general_settings.pass_through_endpoints a config-owned key. The DB reader then got the config list back as if it were DB rows, re-registered each entry without forward_headers on every DB sync, and the stripped copy won the route lookup, so a config pass-through with forward_headers: true stopped forwarding Authorization. UI create, update and delete of pass-throughs were also rejected while the config declared any.
This puts pass-throughs back on their pre-#41779 path: the settings store no longer lets the config own the key, the config list is captured env-resolved at load_config, each DB sync merges DB entries with config entries on paths the DB does not declare, and /config/field/info reads the stored rows only. A UI pass-through write re-applies that merge immediately so the config entries stay served until the next sync.
* fix(proxy): keep config pass-throughs in every reload of the merged list
get_config now returns DB pass-throughs plus config ones on other paths,
each DB sync republishes that merged list, and /config/field/info reads
pass_through_endpoints from the DB row so a UI write never drops stored
entries when models are not stored in the DB
* fix(proxy): keep serving pass-throughs while the config file reloads
load_yaml cleared the runtime pass-through list, so auth: false routes
answered 401 while get_config awaited the database
* fix(proxy): read stored pass-throughs from the writer before a UI write
A lagging read replica could return an older list, and the UI create and
edit flows write the whole field back
* fix(proxy): apply config file pass-through auth changes on reload
The kept runtime list was merged as if it were DB entries, so an edited
config entry on the same path was dropped. Merge the stored DB row with
the fresh config instead, and give the field-info test mock a writer
* fix(proxy): keep pass-throughs served while a DB sync reads the database
get_config resets the stored DB rows before reading them again, which
cleared the served pass-through list and made auth: false routes answer
401 for the length of the read
* refactor(proxy): move the settings store reload out of the loop
basedpyright rejects a Final variable assigned inside a loop
(cherry picked from commit 2eb2bf130b)
* fix(proxy): unregister logging callbacks removed from the stored config
POST /config/callback/delete saved the config and resynced, but the resync only
ever added callbacks, so a deleted callback kept exporting and kept showing in
/get/config/callbacks as read-only on every worker.
ProxyConfig now tracks which callback list entries each DB config sync
registered and unregisters them once the stored config stops listing them.
Callbacks it did not register (YAML, code) are never touched, and a failed
config load skips the sync instead of treating the config as empty.
* refactor(proxy): keep callback sync comprehensions to one for clause
* feat(otel): emit gen_ai.conversation.id from the caller's session id on v2 LLM spans (#42486)
* feat(otel): emit gen_ai.conversation.id from the caller's session id on v2 LLM spans
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep the caller's header session under missing_session_id: generate and read replayed payload session ids
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): drop only the proxy-minted session id so a caller id on the other metadata key survives
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep a replayed session id hidden when it only echoes the payload trace id
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep a replayed session id even when the payload trace id fell back to it
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): stop reading the replayed payload's session id, the generated marker does not survive replay
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): audit gen_ai.conversation.id on otel v2 spans through a real proxy, sink and postgres
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): keep otel conversation rigs alive for the whole session so shuffled shards do not reboot the proxy per test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(otel): stop the audit rig proxies from probing sibling test peers for model info
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(otel): record accepted OTLP batches in the sink instead of mutating the collector
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(otel): guard the accepted batch deque so snapshots cannot race sink appends
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: mrinal <mrinal@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: yucheng <yucheng@berri.ai>
(cherry picked from commit a3196907e4)
* fix(jwt): accept a team alias in x-litellm-team-id (#42445)
* fix(jwt): accept a team alias in x-litellm-team-id
The header only matched canonical team ids, so a JWT caller selecting one of their teams by its alias got a 403 even though they belonged to it. The header value is now resolved through the existing alias lookup before the JWT allowed-team check and the DB membership fallback, while a value that is already a team id never costs an alias lookup and denials keep naming the value the caller sent
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(jwt): only alias a header team id the database provably lacks
Under fallback_to_db_teams a header value whose team row read fails for any reason other than TeamNotFoundError now keeps the membership denial instead of falling through to the alias lookup, so a degraded read cannot select a different team that carries the value as an alias. Drops the HeaderTeam docstring that only restated its fields
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: ryan <ryan@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
(cherry picked from commit 071cb49d32)
* fix(jwt): say x-litellm-team-id matched no team id or alias in the 403 (#42495)
* fix(jwt): say x-litellm-team-id matched no team id or alias in the 403
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(jwt): tell the caller when x-litellm-team-id names an alias shared by several teams
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(jwt): deny a shared x-litellm-team-id alias exactly like an unknown value
A distinct 403 for an alias several teams share was raised before the
allowed-teams check, so any JWT could probe which aliases exist. The
alias lookup now treats the duplicate as a miss, and both denials say
the value does not resolve to a team id or a unique team alias, which
is true for unknown, unauthorized and duplicate values alike
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: ryan <ryan@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
(cherry picked from commit 08639fcf42)
* fix(jwt): let x-litellm-team-id select DB membership teams when the token also carries a team claim (#43206)
* fix(jwt): let x-litellm-team-id select DB membership teams when the token also carries a team claim
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* docs(jwt): describe header team selection under fallback_to_db_teams
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: yassin <yassin@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
(cherry picked from commit 7b4fd47c6e)
---------
Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: mrinal <mrinal@berri.ai>
Co-authored-by: yucheng <yucheng@berri.ai>
Co-authored-by: ryan <ryan@berri.ai>
Co-authored-by: yassin <yassin@berri.ai>
* fix: apply configured cache_control_injection_points beside client cache_control marks (#41956)
(cherry picked from commit d8267d507d)
* fix(token_counter): count replayed redacted_thinking blocks so prompt_caching keeps pinning (#42069)
(cherry picked from commit 0e7cf5113e)
* fix(mcp): return camelCase tool keys from /v1/mcp/tools after the SDK 2 upgrade (#42352)
(cherry picked from commit 5eb4e30f29)
* fix(rust_bridge): keep the Messages route on Python until the Rust path is ready (#42517)
rc/1.103.0 has no token counter or tokenizer routes in the catalog, so only the Messages rule moves to PYTHON_ONLY
(cherry picked from commit 85ed18e8f2)
* fix(proxy): stop leaking periodic tasks on every DB config reload (#42784)
(cherry picked from commit fc87a06f00)
---------
Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com>
Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Prisma types a raw array parameter from the first batch a connection sees. After a flush in
which every member cost was a whole number (a free model), the connection's cached statement
expected int8[] and every later fractional batch on it failed with "improper binary format in
array element", so member spend silently stopped landing while team spend kept rising.
The rows now travel as one JSON document unpacked by jsonb_to_recordset with the column types
declared in SQL, so Postgres types the numbers and the batch shape no longer matters.
(cherry picked from commit 5e4b1b9df0)
Co-authored-by: ryan-crabbe-berri <ryan@berri.ai>
Reverts the code from #41324 (merge 87694c26ef) and #41293 (merge 2e46b10320).
The LiteLLM_DailyGlobalSpend model and migration stay so databases that applied
rc.1 keep a consistent table and rollup marker when the feature returns.
The stored flag is the cluster's desired state and is what the form saves, so it
cannot also stand as proof that this process activated the callback: a pod that
lagged or failed to apply it would still read as on. Report the process's own
registration as a separate read-only field and warn on the page when the two
disagree, so a failed activation is visible instead of only logged.
Deriving enabled from whether this process has the callback registered makes a
pod that has not polled yet report off while the cluster runs it, and the next
save writes that off back for every pod. The stored flag is the cluster's own
answer, so prefer it and fall back to local registration only when none is
stored, which is the config-activated case that has no flag to read.
The member add, delete and role-change audit rows were awaited on the
request path, so a slow audit sink held the response, and the roster was
serialized before checking whether audit logging is enabled at all.
Membership audit work is now scheduled after the transaction commits and
skipped outright when auditing is off.
member_update read the roster outside the team advisory lock and wrote
it back, so a concurrent add or delete could be lost. It now takes the
lock, rereads the roster, and builds the before and after snapshots from
that read.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
The last-known org copy was written only on get_org_object's DB-read path. The
virtual-key auth prefetch fills the same 5s org entry directly, so with keys and
JWTs of one org on the same worker the JWT lookup always hit the cache, never
wrote the copy, and a DB outage turned that JWT traffic into 503s again.
get_org_object_for_request now writes the copy itself whenever this worker holds
none, under the management-object TTL, and get_org_object is back to its shape
on main.
An explicit target_storage=litellm_db upload was accepted for any model, so an OpenAI model's litellm_db://<uuid> id was sent to OpenAI as input_file_id and a model-less upload left a content row nothing can read; it now answers 400 on target_storage. An explicit target_storage skips the files api probe and the purpose and single-target gates, which only decide whether LiteLLM keeps the file itself, so an azure_storage user_data upload for a vLLM model reaches the storage path again as it did before this branch. cancel_batch authorizes the model of every LiteLLM-managed batch id before it branches, the way retrieve_batch already does, so the LiteLLM-executed branch gets the check its provider sibling had. Restores the test_afile_delete_passes_trusted_model_credentials_to_router definition line an earlier commit dropped
Peer pods gate the settings poll on general_settings.supported_db_objects,
which validates against SupportedDBObjectType. Without a member for this
name an operator could not opt in, so a configured allowlist left every
pod but the one that served the write on stale settings.
Also types the dashboard's settings payload off the generated schema
instead of Record<string, any>.
Greptile flagged the unannotated list and append against the repository's
immutable-state and Final-local rules (LIT001/LIT010). Recording the call on an
AsyncMock removes the accumulator entirely and matches how the neighbouring
audit-log tests in this file read their captured arguments.
A JWT whose team sits in an org resolves the org on every request, and the
org row is cached for only DEFAULT_IN_MEMORY_TTL seconds while the team and
user rows ride the 60s management-object TTL. A few seconds into a database
outage the org lookup failed closed and that traffic got 503s while the same
request through a virtual key kept succeeding on its cached team.
get_org_object now also keeps a last-known copy of the org row under the
management-object TTL, and get_org_object_for_request serves that copy when
the database is unreachable, so JWT traffic degrades the same way the team
lookup does. A missing copy keeps the previous behaviour: fail closed unless
allow_requests_on_db_unavailable is set.
Two tests in the config/misc management suite were failing every run against
the Buildkite e2e stack, and one of them took the rest of the build with it.
test_add_allowed_ip_does_not_store_unrelated_config_value posted 127.0.0.1 to
/add/allowed_ip. That route sets the live general_settings["allowed_ips"] that
auth_utils._check_valid_ip reads before it persists anything, and the check is
exact string membership with no CIDR support, so from the moment the POST
returns only 127.0.0.1 can reach the proxy. The runner 403s on its very next
call, and the deferred /delete/allowed_ip sits behind the same auth dependency,
so the cleanup is locked out too and every later test in the build 403s. Build
254's first attempt lost 459 of its 465 failures to that one cascade.
There is no safe way to exercise the route against a shared proxy: nothing
reports the caller's address as the proxy sees it, so a test cannot allowlist
itself first. Move the claim to the route's own TestClient suite, where the
auth dependency is overridden and general_settings is per-test, and record the
route in the module docstring beside /cache/settings and the Vault override so
it is not re-added. save_config's end of the contract was already covered by
test_ProxyConfig_save_config_merges_changed_keys_without_copying_file_settings;
the new test covers the route's end, that what it hands save_config differs
from the loaded config in allowed_ips and nothing else.
The unrelated-key probe also only ever worked on one lane: max_parallel_requests
was added to tests/e2e/gateway/stage_mirror_ci_config.yml and never to the
Buildkite stack's config, where resolve() reports it as "unset" rather than
"config". That key is now unused, so drop it again.
test_config_update_persists_router_setting_to_get wrote router_settings.
num_retries, which both lanes declare in their config file, so the config-
ownership work correctly refuses it with a 400. Switch to retry_after, which is
declared by neither lane, is accepted by /config/update, and is reported back by
GET /router/settings. Verified against a live proxy: max_fallbacks also takes
the write but never reads back, so the read-back poll is what picks the key.
A batch upload naming a model on a LiteLLM-executed provider now checks
that the key may call that model before the upstream server is probed for
a Files API, matching the order batch create already uses. Only targets on
an executed provider are checked here, so provider-model uploads keep
their existing behavior.
File content reads and writes move out of the storage backend into
ManagedFileContentRepository, so the backend no longer queries Prisma
directly.
