- 18 Mar, 2024 19 commits
-
-
Dridi Boukelmoune authored
When a stream times out waiting for window credits, and all the other streams are broke, declare the whole connection bankrupt.
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
It replaces idle_send_timeout when the stream is waiting for window credits before sending more DATA frames. The idle_send_timeout will still apply to individual writes to the socket, but triggering it is considered a failure condition (this was already the case). The two loops are merged into a single one to better deal with the lack of ordering guarantees of request vs connection stream control flow window crediting. Refs #2980
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
This parameter needs to be somewhat high because web browsers may for example only credit streams for image resources just enough to parse metadata like width and height of a picture in order to perform layout computations and fetch resources more sensitive to latency before the effective image payloads.
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
The new meaning for h2->sess->t_idle is now the time since the last complete frame, except before the first frame where it corresponds to the creation of the session.
-
Dridi Boukelmoune authored
And after issuing or receiving a goaway, stop processing frames when there are no streams left. Initially submitted as part of listen sockets management. Refs #3959
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
Reverting "retrieving" to "processing" is motivated by the fact that we process ESI on the client side and we don't know whether we are in fact retrieving an object from a backend, or just reusing one from the cache. Even if the word "processing" appears again in the sentence, I think it is more accurate here. Better diff with the --word-diff option.
-
Nils Goroll authored
-
Nils Goroll authored
- before coffee + after coffee
-
Nils Goroll authored
-
Dridi Boukelmoune authored
Spotted by Nils.
-
Dridi Boukelmoune authored
All paths to EXP_Remove() raise the OC_F_DYING flag to dismiss the objcore in HSH_Lookup(). A regular purge used to EXP_Rearm() an oc for immediate (yet delayed) expiry, leaving it to the expiry thread to apply this flag and call EXP_Remove(). When the regular purge went from EXP_Rearm() to EXP_Remove() in order to avoid counting purged objects as expired, the OC_F_DYING flag was overlooked. Refs 709f71ab
-
- 15 Mar, 2024 9 commits
-
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Simon Stridsberg authored
Ubuntu noble tries to use `fchmodat2` (new syscall) and gets permission denied instead of ENOSYS. This is a small security risk but it's running inside of circleci containers anyway so i think its acceptable.
-
AlveElde authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
See previous commit.
-
Dridi Boukelmoune authored
But spread the actual changes in their respective sections.
-
- 14 Mar, 2024 1 commit
-
-
Dridi Boukelmoune authored
Quick attempt to address 3f69805f. The remaining TODO in the changes needs to be addressed, and upgrade notes are empty.
-
- 11 Mar, 2024 1 commit
-
-
Walid Boudebouda authored
-
- 07 Mar, 2024 4 commits
-
-
Nils Goroll authored
weird behavior was observed while working on a new feature: This test would terminate s1 after the delay before varnish could finish reading all of the chunked body, which would leave an unfinished backend request in varnishd when it terminated.
-
Asad Sajjad Ahmed authored
Signed-off-by: Asad Sajjad Ahmed <asadsa@varnish-software.com>
-
Nils Goroll authored
Mention the old/new cli_limit values Polish wordning of bits parameter change Add param.show example in place of what looks like an unmotivated param.set -j example
-
Nils Goroll authored
-
- 06 Mar, 2024 2 commits
-
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
- 05 Mar, 2024 2 commits
-
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
Please note the large TODO regarding the rapid reset attack mitigation.
-
- 25 Sep, 2018 1 commit
-
-
Geoff Simmons authored
Restructured so that: * 'Upgrading' is limited to work that has to be done to upgrade from a current deployment to the new version. * 'Changes' is a comprehensive, user-level description of changes and new features. Conflicts: doc/sphinx/whats-new/changes-trunk.rst doc/sphinx/whats-new/index.rst doc/sphinx/whats-new/upgrading-trunk.rst
-
- 05 Mar, 2024 1 commit
-
-
Dridi Boukelmoune authored
Soptted by Stéphane Cance.
-