You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@trafficserver.apache.org by "masaori335 (via GitHub)" <gi...@apache.org> on 2026/05/13 07:50:43 UTC

[PR] Fix RangeTransform on stale-revalidate [trafficserver]

masaori335 opened a new pull request, #13158:
URL: https://github.com/apache/trafficserver/pull/13158

   Fix RangeTransform on the stale-revalidate path when the fresh origin body is differs from the stale object size.
   
   For example, in the added AuTest, original object size was `64097 ` and updated size was `40000`, the second (revalidated) response from ATS was below without this patch: 
   ```
   HTTP/1.1 206 Partial Content
   Content-Range: bytes 0-64096/40000
   Content-Length: 64097
   ```
   while the actual response body size was 40000 bytes.
   
   To fix the issue, 
   
   1. Update `m_write_vio.ndone` in `RangeTransform::transform_to_range()`.
   2. Re-parse the Range against the fresh `Content-Length` from origin server.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "masaori335 (via GitHub)" <gi...@apache.org>.
masaori335 commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3245072072


##########
src/proxy/http/HttpSM.cc:
##########
@@ -5091,6 +5091,35 @@ HttpSM::do_range_setup_if_necessary()
         if (t_state.cache_info.action == HttpTransact::CacheAction_t::REPLACE) {
           if (t_state.hdr_info.server_response.status_get() == HTTPStatus::OK) {
             Dbg(dbg_ctl_http_range, "Serving transform after stale cache re-serve");
+
+            // Ranges and range_output_cl were computed against the stale cached object size. If the fresh origin Content-Length
+            // differs, re-parse the Range against the fresh value so the outgoing Content-Length/Content-Range match the body
+            // actually being sent. Without this, Content-Length/Content-Range advertise the stale cached size.
+            const int64_t fresh_cl  = t_state.hdr_info.server_response.get_content_length();
+            const int64_t cached_cl = t_state.cache_info.object_read ? t_state.cache_info.object_read->object_size_get() : -1;
+            if (fresh_cl > 0 && fresh_cl != cached_cl) {

Review Comment:
   Make sense. We need to cover the case of new Content-Length is 0.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "masaori335 (via GitHub)" <gi...@apache.org>.
masaori335 commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3245280285


##########
src/proxy/http/HttpSM.cc:
##########
@@ -5091,6 +5091,35 @@ HttpSM::do_range_setup_if_necessary()
         if (t_state.cache_info.action == HttpTransact::CacheAction_t::REPLACE) {
           if (t_state.hdr_info.server_response.status_get() == HTTPStatus::OK) {
             Dbg(dbg_ctl_http_range, "Serving transform after stale cache re-serve");
+
+            // Ranges and range_output_cl were computed against the stale cached object size. If the fresh origin Content-Length
+            // differs, re-parse the Range against the fresh value so the outgoing Content-Length/Content-Range match the body
+            // actually being sent. Without this, Content-Length/Content-Range advertise the stale cached size.
+            const int64_t fresh_cl  = t_state.hdr_info.server_response.get_content_length();
+            const int64_t cached_cl = t_state.cache_info.object_read ? t_state.cache_info.object_read->object_size_get() : -1;
+            if (fresh_cl > 0 && fresh_cl != cached_cl) {

Review Comment:
   Done.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "bneradt (via GitHub)" <gi...@apache.org>.
bneradt commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3262731158


##########
tests/gold_tests/headers/range_transform.test.py:
##########
@@ -0,0 +1,18 @@
+#  Licensed to the Apache Software Foundation (ASF) under one
+#  or more contributor license agreements.  See the NOTICE file
+#  distributed with this work for additional information
+#  regarding copyright ownership.  The ASF licenses this file
+#  to you under the Apache License, Version 2.0 (the
+#  "License"); you may not use this file except in compliance
+#  with the License.  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+tr = Test.ATSReplayTest(replay_file="replays/range_transform.replay.yaml")
+tr.Processes.Default.TimeOut = 10

Review Comment:
   Might not need this TimeOut set if we adjust the delay calls above.



##########
tests/gold_tests/headers/replays/range_transform.replay.yaml:
##########
@@ -0,0 +1,223 @@
+#  Licensed to the Apache Software Foundation (ASF) under one
+#  or more contributor license agreements.  See the NOTICE file
+#  distributed with this work for additional information
+#  regarding copyright ownership.  The ASF licenses this file
+#  to you under the Apache License, Version 2.0 (the
+#  "License"); you may not use this file except in compliance
+#  with the License.  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+# Verify RangeTransform on stale-revalidate
+#
+# Preconditions for the bug:
+#  1. Client request has a `Range` header.
+#  2. Cached object is stale, so ATS revalidates.
+#  3. Origin returns `200 OK` with the full body (not 304, not 206).
+#  4. The fresh `Content-Length` differs from the cached object size.
+
+meta:
+  version: '1.0'
+
+autest:
+  description: 'Verify RangeTransform'
+
+  dns:
+    name: "dns-range-transform"
+
+  server:
+    name: "server-range-transform"
+
+  client:
+    name: "client-range-transform"
+
+  ats:
+    name: "ts-range-transform"
+
+    process_config:
+      enable_cache: true
+
+    records_config:
+      proxy.config.http.wait_for_cache: 1
+      proxy.config.http.cache.required_headers: 0
+      proxy.config.diags.debug.enabled: 1
+      proxy.config.diags.debug.tags: 'http'
+
+    remap_config:
+      - from: "http://example.com/"
+        to: "http://backend.example.com:{SERVER_HTTP_PORT}/"
+
+    log_validation:
+      traffic_out:
+        contains:
+          - expression: 'perform_transform_cache_write_action CacheAction_t::REPLACE'
+            description: 'Stale cache is replaced via RangeTransform'
+
+sessions:
+  # Prime cache with BIG body.
+  - transactions:
+      - client-request:
+          method: "GET"
+          version: "1.1"
+          url: /obj
+          headers:
+            fields:
+              - [Host, example.com]
+              - [uuid, prime]
+
+        server-response:
+          status: 200
+          reason: OK
+          headers:
+            fields:
+              - [Date, "Mon, 01 Jan 2026 00:00:00 GMT"]
+              - [Cache-Control, "max-age=1, public"]
+              - [Content-Type, application/octet-stream]
+              - [Content-Length, 64097]
+          content:
+            size: 64097
+
+        proxy-response:
+          status: 200
+
+  # Stale revalidate: origin body shrunk to 40000, Range end (64096) unreachable.
+  - transactions:
+      - client-request:
+          delay: 1500ms

Review Comment:
   Nit.: 100ms should be plenty for IO. We use that elsewhere in the autests for cache IO. That will save a few seconds of test execution here.
   
   This applies to quite a few `delay:` calls in here.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "masaori335 (via GitHub)" <gi...@apache.org>.
masaori335 commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3245281579


##########
src/proxy/http/HttpSM.cc:
##########
@@ -5091,6 +5091,35 @@ HttpSM::do_range_setup_if_necessary()
         if (t_state.cache_info.action == HttpTransact::CacheAction_t::REPLACE) {
           if (t_state.hdr_info.server_response.status_get() == HTTPStatus::OK) {
             Dbg(dbg_ctl_http_range, "Serving transform after stale cache re-serve");
+
+            // Ranges and range_output_cl were computed against the stale cached object size. If the fresh origin Content-Length
+            // differs, re-parse the Range against the fresh value so the outgoing Content-Length/Content-Range match the body
+            // actually being sent. Without this, Content-Length/Content-Range advertise the stale cached size.
+            const int64_t fresh_cl  = t_state.hdr_info.server_response.get_content_length();
+            const int64_t cached_cl = t_state.cache_info.object_read ? t_state.cache_info.object_read->object_size_get() : -1;
+            if (fresh_cl > 0 && fresh_cl != cached_cl) {
+              SMDbg(dbg_ctl_http_range, "Re-parsing range against fresh origin Content-Length %" PRId64 " (was %" PRId64 ")",
+                    fresh_cl, cached_cl);
+              delete[] t_state.ranges;
+              t_state.ranges           = nullptr;
+              t_state.num_range_fields = 0;
+              t_state.range_setup      = HttpTransact::RangeSetup_t::NONE;
+              t_state.range_output_cl  = 0;
+              parse_range_done         = false;
+
+              std::string_view content_type =
+                t_state.hdr_info.server_response.value_get(static_cast<std::string_view>(MIME_FIELD_CONTENT_TYPE));
+              parse_range_and_compare(field, fresh_cl);
+              calculate_output_cl(content_type.length(), num_chars_for_int(fresh_cl));
+
+              if (t_state.range_setup != HttpTransact::RangeSetup_t::REQUESTED) {
+                // Re-parse yielded e.g. RANGE_NOT_SATISFIABLE (entire range
+                // past fresh body); let downstream handling take over without
+                // installing the transform.

Review Comment:
   Add a test cases in AuTest to cover this scenario.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "cmcfarlen (via GitHub)" <gi...@apache.org>.
cmcfarlen commented on PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#issuecomment-4555738933

   Cherry-picked to 10.2.x


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "bneradt (via GitHub)" <gi...@apache.org>.
bneradt commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3262791362


##########
tests/gold_tests/headers/replays/range_transform.replay.yaml:
##########
@@ -0,0 +1,223 @@
+#  Licensed to the Apache Software Foundation (ASF) under one
+#  or more contributor license agreements.  See the NOTICE file
+#  distributed with this work for additional information
+#  regarding copyright ownership.  The ASF licenses this file
+#  to you under the Apache License, Version 2.0 (the
+#  "License"); you may not use this file except in compliance
+#  with the License.  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+# Verify RangeTransform on stale-revalidate
+#
+# Preconditions for the bug:
+#  1. Client request has a `Range` header.
+#  2. Cached object is stale, so ATS revalidates.
+#  3. Origin returns `200 OK` with the full body (not 304, not 206).
+#  4. The fresh `Content-Length` differs from the cached object size.
+
+meta:
+  version: '1.0'
+
+autest:
+  description: 'Verify RangeTransform'
+
+  dns:
+    name: "dns-range-transform"
+
+  server:
+    name: "server-range-transform"
+
+  client:
+    name: "client-range-transform"
+
+  ats:
+    name: "ts-range-transform"
+
+    process_config:
+      enable_cache: true
+
+    records_config:
+      proxy.config.http.wait_for_cache: 1
+      proxy.config.http.cache.required_headers: 0
+      proxy.config.diags.debug.enabled: 1
+      proxy.config.diags.debug.tags: 'http'
+
+    remap_config:
+      - from: "http://example.com/"
+        to: "http://backend.example.com:{SERVER_HTTP_PORT}/"
+
+    log_validation:
+      traffic_out:
+        contains:
+          - expression: 'perform_transform_cache_write_action CacheAction_t::REPLACE'
+            description: 'Stale cache is replaced via RangeTransform'
+
+sessions:
+  # Prime cache with BIG body.
+  - transactions:
+      - client-request:
+          method: "GET"
+          version: "1.1"
+          url: /obj
+          headers:
+            fields:
+              - [Host, example.com]
+              - [uuid, prime]
+
+        server-response:
+          status: 200
+          reason: OK
+          headers:
+            fields:
+              - [Date, "Mon, 01 Jan 2026 00:00:00 GMT"]
+              - [Cache-Control, "max-age=1, public"]
+              - [Content-Type, application/octet-stream]
+              - [Content-Length, 64097]
+          content:
+            size: 64097
+
+        proxy-response:
+          status: 200
+
+  # Stale revalidate: origin body shrunk to 40000, Range end (64096) unreachable.
+  - transactions:
+      - client-request:
+          delay: 1500ms

Review Comment:
   Oh, this needs to exceed the 1s max age. I missed that. nvm, what you have makes sense then.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "masaori335 (via GitHub)" <gi...@apache.org>.
masaori335 merged PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "Copilot (via GitHub)" <gi...@apache.org>.
Copilot commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3243715383


##########
tests/gold_tests/headers/range_transform.test.py:
##########
@@ -0,0 +1,18 @@
+#  Licensed to the Apache Software Foundation (ASF) under one
+#  or more contributor license agreements.  See the NOTICE file
+#  distributed with this work for additional information
+#  regarding copyright ownership.  The ASF licenses this file
+#  to you under the Apache License, Version 2.0 (the
+#  "License"); you may not use this file except in compliance
+#  with the License.  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+

Review Comment:
   This new gold test doesn’t follow the common `headers/*.test.py` pattern of providing a module docstring and setting `Test.Summary` (and usually `Test.ContinueOnFail = True`). Adding those improves test reporting consistency (see e.g. tests/gold_tests/headers/range.test.py).
   



##########
src/proxy/http/HttpSM.cc:
##########
@@ -5091,6 +5091,35 @@ HttpSM::do_range_setup_if_necessary()
         if (t_state.cache_info.action == HttpTransact::CacheAction_t::REPLACE) {
           if (t_state.hdr_info.server_response.status_get() == HTTPStatus::OK) {
             Dbg(dbg_ctl_http_range, "Serving transform after stale cache re-serve");
+
+            // Ranges and range_output_cl were computed against the stale cached object size. If the fresh origin Content-Length
+            // differs, re-parse the Range against the fresh value so the outgoing Content-Length/Content-Range match the body
+            // actually being sent. Without this, Content-Length/Content-Range advertise the stale cached size.
+            const int64_t fresh_cl  = t_state.hdr_info.server_response.get_content_length();
+            const int64_t cached_cl = t_state.cache_info.object_read ? t_state.cache_info.object_read->object_size_get() : -1;
+            if (fresh_cl > 0 && fresh_cl != cached_cl) {
+              SMDbg(dbg_ctl_http_range, "Re-parsing range against fresh origin Content-Length %" PRId64 " (was %" PRId64 ")",
+                    fresh_cl, cached_cl);
+              delete[] t_state.ranges;
+              t_state.ranges           = nullptr;
+              t_state.num_range_fields = 0;
+              t_state.range_setup      = HttpTransact::RangeSetup_t::NONE;
+              t_state.range_output_cl  = 0;
+              parse_range_done         = false;
+
+              std::string_view content_type =
+                t_state.hdr_info.server_response.value_get(static_cast<std::string_view>(MIME_FIELD_CONTENT_TYPE));
+              parse_range_and_compare(field, fresh_cl);
+              calculate_output_cl(content_type.length(), num_chars_for_int(fresh_cl));
+
+              if (t_state.range_setup != HttpTransact::RangeSetup_t::REQUESTED) {
+                // Re-parse yielded e.g. RANGE_NOT_SATISFIABLE (entire range
+                // past fresh body); let downstream handling take over without
+                // installing the transform.

Review Comment:
   In the stale-revalidate re-parse block, returning early when the new parse result is NOT REQUESTED leaves the transaction without a transform hook and without building a 416 (or other) response for an unsatisfiable Range. In the REPLACE path, HttpTransact does not check range_setup after do_range_setup_if_necessary(), so this can silently fall back to serving the full 200 response to a Range request. Consider explicitly handling NOT_SATISFIABLE / NOT_HANDLED here (e.g., convert to a 416, or force the transaction onto an error path) rather than just returning.
   



##########
src/proxy/http/HttpSM.cc:
##########
@@ -5091,6 +5091,35 @@ HttpSM::do_range_setup_if_necessary()
         if (t_state.cache_info.action == HttpTransact::CacheAction_t::REPLACE) {
           if (t_state.hdr_info.server_response.status_get() == HTTPStatus::OK) {
             Dbg(dbg_ctl_http_range, "Serving transform after stale cache re-serve");
+
+            // Ranges and range_output_cl were computed against the stale cached object size. If the fresh origin Content-Length
+            // differs, re-parse the Range against the fresh value so the outgoing Content-Length/Content-Range match the body
+            // actually being sent. Without this, Content-Length/Content-Range advertise the stale cached size.
+            const int64_t fresh_cl  = t_state.hdr_info.server_response.get_content_length();
+            const int64_t cached_cl = t_state.cache_info.object_read ? t_state.cache_info.object_read->object_size_get() : -1;
+            if (fresh_cl > 0 && fresh_cl != cached_cl) {

Review Comment:
   The re-parse is guarded by `fresh_cl > 0`, which skips the correction when the origin returns `Content-Length: 0` (valid for an empty 200 response). In that case, range_output_cl/ranges remain based on the stale cached size, and the transform path can generate inconsistent headers/body again. Consider triggering the re-parse for `fresh_cl >= 0` when Content-Length is present (and handling the `0` case appropriately, likely as NOT_SATISFIABLE for any non-empty Range).
   



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] Fix RangeTransform on stale-revalidate [trafficserver]

Posted by "masaori335 (via GitHub)" <gi...@apache.org>.
masaori335 commented on code in PR #13158:
URL: https://github.com/apache/trafficserver/pull/13158#discussion_r3245236190


##########
src/proxy/http/HttpSM.cc:
##########
@@ -5091,6 +5091,35 @@ HttpSM::do_range_setup_if_necessary()
         if (t_state.cache_info.action == HttpTransact::CacheAction_t::REPLACE) {
           if (t_state.hdr_info.server_response.status_get() == HTTPStatus::OK) {
             Dbg(dbg_ctl_http_range, "Serving transform after stale cache re-serve");
+
+            // Ranges and range_output_cl were computed against the stale cached object size. If the fresh origin Content-Length
+            // differs, re-parse the Range against the fresh value so the outgoing Content-Length/Content-Range match the body
+            // actually being sent. Without this, Content-Length/Content-Range advertise the stale cached size.
+            const int64_t fresh_cl  = t_state.hdr_info.server_response.get_content_length();
+            const int64_t cached_cl = t_state.cache_info.object_read ? t_state.cache_info.object_read->object_size_get() : -1;
+            if (fresh_cl > 0 && fresh_cl != cached_cl) {
+              SMDbg(dbg_ctl_http_range, "Re-parsing range against fresh origin Content-Length %" PRId64 " (was %" PRId64 ")",
+                    fresh_cl, cached_cl);
+              delete[] t_state.ranges;
+              t_state.ranges           = nullptr;
+              t_state.num_range_fields = 0;
+              t_state.range_setup      = HttpTransact::RangeSetup_t::NONE;
+              t_state.range_output_cl  = 0;
+              parse_range_done         = false;
+
+              std::string_view content_type =
+                t_state.hdr_info.server_response.value_get(static_cast<std::string_view>(MIME_FIELD_CONTENT_TYPE));
+              parse_range_and_compare(field, fresh_cl);
+              calculate_output_cl(content_type.length(), num_chars_for_int(fresh_cl));
+
+              if (t_state.range_setup != HttpTransact::RangeSetup_t::REQUESTED) {
+                // Re-parse yielded e.g. RANGE_NOT_SATISFIABLE (entire range
+                // past fresh body); let downstream handling take over without
+                // installing the transform.

Review Comment:
   Returning 416 is another choice for this case, however as [RFC 9110 15.5.16 416 Range Not Satisfiable](https://httpwg.org/specs/rfc9110.html#status.416) says servers are free to ignore Range and return 200 OK. This align with current our behavior.
   
   > Note: Because servers are free to ignore [Range](https://httpwg.org/specs/rfc9110.html#field.range), many implementations will respond with the entire selected representation in a [200 (OK)](https://httpwg.org/specs/rfc9110.html#status.200) response.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@trafficserver.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org