- 31 Oct, 2019 5 commits
-
-
Poul-Henning Kamp authored
from both client and backend side.
-
Dag Haavi Finstad authored
-
Nils Goroll authored
When we breed threads without a delay, we should make an effort to let them run as soon as possible and not have the cpu occupied with the herder code. This will increase the overhead due to additional context switching for the benefit of reducing latencies until new threads get to run. Effect seen with this vtc by @Dridi: varnishtest "over-breeding" varnish v1 -arg "-p thread_pools=1 -p thread_pool_min=10" varnish v1 -vcl { backend be none; } -start varnish v1 -expect MAIN.threads == 10 Tested with the following script: ./varnishtest -i -k -j40 -n1000 tests/as_above.vtc | awk '/^#/ { print $5; }'| sort | uniq -c Multiple runs show a consistent drop of failed tests: before | after -------+------ 76 | 18 67 | 24 66 | 18
-
Poul-Henning Kamp authored
Fixes #3045
-
Federico G. Schwindt authored
-
- 30 Oct, 2019 10 commits
-
-
Nils Goroll authored
-
Nils Goroll authored
Found by @Dridi
-
Nils Goroll authored
I had previously overlooked that there exists a more complete regression test by him in #3010 Integrate that vtc, polished and modified to account for the backend 503 for no restart after rollback, into m17, making r3009 obsolete again
-
Nils Goroll authored
Fixes #3083 to get there, also centralize req->resp setup for deliver and synth: No semantic changes other than some reordering, which also fixes an odd log line ordering as shown by the change to c00018.vtc
-
Dridi Boukelmoune authored
The same can be achieved with `varnishtest -p debug=+witness`, already used by `make witness`.
-
Dridi Boukelmoune authored
This change introduces a top-level make witness target that builds a dot graph and if graphviz is available, an SVG file as well. A shell script replaces the previous python script that no longer works. Instead of fixing witness.py, which is probably trivial, the shell script does an intermediate pass and programmatically looks for cycles using tsort(1). Checking lock dependencies becomes actionable in a CI context. The script also takes explicit test directories on purpose, to have the ability to aggregate test results from multiple executions. For example when the test suite is run on various operating systems or with varying privileges to cover feature-conditional tests.
-
Dridi Boukelmoune authored
The simple fact that Witness records might show up in the log might break logexpect commands. There's no reason why we'd want to expect Witness records since their purpose is to be checked after the test finishes.
-
Nils Goroll authored
The gunzip vdp failed to handle junk after end of gzip data. This basically mirrors #942 on the client side, also the fix is basically the same as 41f7a356 The impact of this bug is likely to be low, because the built-in beresp.filters logic will push the testgunzip VFP for gzip content received from backends, so, unless VCL is forced to pass backend responses unchecked or vmods generate body data, it can be considered unlikely that this issue will be hit. Fixes #3109
-
Nils Goroll authored
We use the first 64bit of a sha256 as our pool id, assuming that those are safe enough against collisions. Ensure we do not fail on that assumption. This also makes it a caller error to deliberately use the same pool id for different endpoints. As this was not possible before the id change, I do not consider it a regression. Ref: ec70dbc7
-
Nils Goroll authored
Ref: ec70dbc7 struct vtp_cs was only used for comparisons during pool lookup
-
- 29 Oct, 2019 5 commits
-
-
Dridi Boukelmoune authored
If varnish closes the connection while a client is waiting for an rxsomething command you may run into this kind of scenario: *** c1 HTTP2 rx failed (fd:20 read: Connection reset by peer) *** c1 rx: stream: 0, type: DATA (0), flags: 0x00, size: 0 **** c1 s0 - no data ---- c1 Wrong frame type DATA (0) wanted SETTINGS This was already done correctly for the frame body.
-
Dridi Boukelmoune authored
Instead of having to manually document them, we can now let the RST dump figure everything out. Refs #3099
-
Dridi Boukelmoune authored
Instead, they are added where they are needed: only in error messages involving them. Refs #3099
-
Dridi Boukelmoune authored
Refs #3108
-
Nils Goroll authored
* clarify defaults * document Transient Fixes #3108
-
- 28 Oct, 2019 1 commit
-
-
Dag Haavi Finstad authored
Fixes: #3086
-
- 25 Oct, 2019 1 commit
-
-
Dridi Boukelmoune authored
Refs #2872
-
- 24 Oct, 2019 5 commits
-
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
Otherwise it is considered a code block instead of a paragraph where it is used.
-
Dridi Boukelmoune authored
The VSL query could in theory do that, but I don't think it was actionable without double quotes to delimit empty fields or fields containing spaces.
-
Dridi Boukelmoune authored
The test case covers the last field of Backend_health records.
-
Dridi Boukelmoune authored
When we output a VSL field containing spaces, all bets are off for VSL queries relying on that field or subsequent fields in the same record. The solution is to allow a quoted-string format for such fields.
-
- 23 Oct, 2019 1 commit
-
-
Guillaume Quintard authored
- fix centos:7 distcheck - use centos to build the dist tarball - simplify git cloning with builtin command - no need for ssh when building - tar the dsc files with the deb ones - use the regular pkg repo now that the alpine PR has been merged
-
- 22 Oct, 2019 1 commit
-
-
Dridi Boukelmoune authored
-
- 21 Oct, 2019 11 commits
-
-
Nils Goroll authored
-
Poul-Henning Kamp authored
-
Nils Goroll authored
brought up on IRC by @scn
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
It is only really needed for one last use case.
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
Inform users when the parameter we try to change depends on another parameter. This is not meant to be merged as-is, we could have a new VSB function to trim the end of the string if it matches the one passed as the second argument. In this case, we would trim the trailing newline only of the CLI response didn't overflow in the first place, adding the extra information on the same line. Alternatively we could also simply add the explanation for the dynamic bounds on the next line. Feedback welcome. Fixes #3098
-
Dridi Boukelmoune authored
-