| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
I missed a case.
|
| |
|
|
|
|
|
|
|
|
|
| |
* Fix quoting so that it works with all OTP versions 20 through 24 [1].
* Dist API in 20 [2] did not have a `listen/2` [3] callback. Implement
`listen/1` so we're compatible with all the supported OTP versions.
[1] https://github.com/apache/couchdb/issues/3821#issuecomment-985089867
[2] https://github.com/erlang/otp/blob/maint-20/lib/kernel/src/inet_tcp_dist.erl#L71-L72
[3] https://github.com/erlang/otp/blob/master/lib/kernel/src/inet_tcp_dist.erl#L79-L80
|
| |\
| |
| | |
improve erlang_ls.config
|
| |/ |
|
| | |
|
| |\
| |
| | |
Delete session cookie samesite 3.x
|
| |/ |
|
| | |
|
| |
|
|
|
|
|
|
| |
This flow was presumably in place for some cluster upgrade scenarios
several years ago, and no one ever completed the work to finalize the
event flow. All the ddoc_cache_opener process does at this point is
forward the event back to ddoc_cache_lru, so let's just send that
message directly and remove the superfluous gen_server.
|
| |
|
|
|
|
|
|
|
|
|
| |
Add `static` modifiers to some of the newer functions.
Fix function protype for get_collator to be `void`. This squashes this warning
on MS compiler:
```
'get_collator': no function prototype given: converting '()' to '(void)'
```
|
| |
|
|
|
|
|
| |
Should fix:
```
Warning: A secret was passed to "mail" using Groovy String interpolation, which is insecure.
```
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These exceptions from main were ported over to 3.x
```
--- a/src/chttpd/src/chttpd.erl
+++ b/src/chttpd/src/chttpd.erl
@@ -491,6 +491,7 @@ extract_cookie(#httpd{mochi_req = MochiReq}) ->
end.
%%% end hack
+%% erlfmt-ignore
set_auth_handlers() ->
AuthenticationDefault = "{chttpd_auth, cookie_authentication_handler},
```
```
--- a/src/couch/src/couch_debug.erl
+++ b/src/couch/src/couch_debug.erl
@@ -49,6 +49,7 @@ help() ->
].
-spec help(Function :: atom()) -> ok.
+%% erlfmt-ignore
help(opened_files) ->
```
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
From PR: #3568
A few changes to the formatting logic compared to `main`:
* rebar3 and `erlfmt` don't work with Erlang 20, so added an OTP version
check to skip formatting for Erlang 20
* `make erlfmt-*` commands are faster. Instead of spawning an `erlfmt`
process per-file, spawn one per `directory/*.erl` pattern.
* Remove the non-0 return code work-around. `erlfmt` was returning an error
on `main` because the line width option used for formatting was different
than the one used during format checking. Making both the default removed
all the spurious non-0 exits with the current version of `erlfmt`. Thanks
to Adam Kocoloski for pointing out the issue.
|
| |
|
|
|
|
|
|
|
| |
The `skip.elixir` file was ported [1] from FDB-based main branch but
contains more exclusions than necessary for 3.x.
This removes tests from `skip.elixir` that pass on 3.x.
[1] https://github.com/apache/couchdb/commit/65fbc6cfdb29bab1072a42f4d6005da452a7fa65#diff-d9485d6f3162015017bc09c9db3f87df0ca489d961c44347241a035854eb8044
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Eliminates the following:
warning: variable "epoch" is unused (if the variable is not meant to be used, prefix it with an underscore)
src/couchdb/test/elixir/test/replication_test.exs:1756: ReplicationTest.seq_to_shards/1
warning: variable "uuid" is unused (if the variable is not meant to be used, prefix it with an underscore)
src/couchdb/test/elixir/test/replication_test.exs:1756: ReplicationTest.seq_to_shards/1
warning: variable "string" is unused (if the variable is not meant to be used, prefix it with an underscore)
src/couchdb/test/elixir/test/utf8_test.exs:32: UTF8Test."test UTF8 support"/1
warning: variable "repeated_view_offset_test_fun" does not exist and is being expanded to "repeated_view_offset_test_fun()", please use parentheses to remove the ambiguity or change the variable name
src/couchdb/test/elixir/test/view_offsets_test.exs:63: ViewOffsetTest."test repeated view offsets"/1
|
| |
|
|
| |
Bad Request
|
| |
|
|
|
|
|
|
|
|
| |
The arity 0 version does not exist.
Fixes dialyzer warning:
{warn_callgraph,{"src/weatherreport_node.erl",126},
{call_to_missing,[weatherreport_check_nodes_connected,check,
0]}}
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Previously, view reduce collation with keys relied on the keys in the
rows returned from the view shards to exactly match (=:=) the keys
specified in the args. However, in the case when there are multiple
rows which compare equal with the unicode collator, that may not
always be the case.
In that case when the rows are fetched from the row dict by key, they
should be matched using the same collation algorithm as the one used
on the view shards.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The appropriately named `hack_local_ini` function in dev/run lives up
to its name by adding a random `secret` to the end of whichever
section contains the line `"; require_valid_user = false\n"`.
Recent commits [1] updated the `setup` application from
`couch_httpd_auth` to `chttpd_auth` without also changing this line
in local.ini, which had the fun effect of causing `setup` to fail with
this error:
```
[error] 2021-10-30T18:50:15.019274Z node1@127.0.0.1 <0.817.0> 85d0421b9c setup sync_admin results [{badrpc,{'EXIT',{function_clause,[{config,set,["chttpd_auth","secret",undefined,true,nil],[{file,"src/config.erl"},{line,189}]},{rpc,'-handle_call_call/6-fun-0-',5,[{file,"rpc.erl"},{line,197}]}]}}},{badrpc,{'EXIT',{function_clause,[{config,set,["chttpd_auth","secret",undefined,true,nil],[{file,"src/config.erl"},{line,189}]},{rpc,'-handle_call_call/6-fun-0-',5,[{file,"rpc.erl"},{line,197}]}]}}}] errors []
[notice] 2021-10-30T18:50:15.019530Z node1@127.0.0.1 <0.817.0> 85d0421b9c 127.0.0.1:15984 127.0.0.1 adm POST /_cluster_setup 500 ok 12
```
This change "fixes" dev/run by also updating the straggling section
name in local.ini.
[1] https://github.com/apache/couchdb/pull/3808
|
| |
|
|
| |
Closes #3801
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, dbs with N < cluster default N would pollute logs with critical
errors regarding not having enough shards. Instead, use each database's
expected N value to emit custodian reports.
Note: the expected N value is a bit tricky to understand since with shard
splitting feature, shard ranges are not guaranteed to exactly match for all
copies. The N value is then defined as the max number of rings which can be
completed with the given set of shards -- complete the ring once, remove
participating shards, try again, etc. Lucky for us, that function is already
written (`mem3_util:calculate_max_n(Shards)` so we are just re-using it.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Prevent failures like:
mem3_rep: find_source_seq_unknown_node_test...*failed*
in function gen_server:call/2 (gen_server.erl, line 206)
in call from couch_log:log/3 (src/couch_log.erl, line 73)
in call from mem3_rep:find_source_seq_int/5 (src/mem3_rep.erl, line 248)
in call from mem3_rep:'-find_source_seq_unknown_node_test/0-fun-0-'/0 (src/mem3_rep.erl, line 794)
**exit:{noproc,{gen_server,call,
[couch_log_server,
{log,{log_entry,warning,<0.17426.5>,
["mem3_rep",32,102,105,110,100|...],
"--------",
["2021",45,"10",45|...]}}]}}
output:<<"">>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Prevent failures like this from repeated test runs:
mem3_bdu_test:73: mem3_bdu_shard_doc_test_ (t_design_docs_are_not_validated)...*failed*
in function mem3_bdu_test:'-t_design_docs_are_not_validated/1-fun-0-'/1 (test/eunit/mem3_bdu_test.erl, line 206)
in call from mem3_bdu_test:t_design_docs_are_not_validated/1 (test/eunit/mem3_bdu_test.erl, line 206)
in call from eunit_test:run_testfun/1 (eunit_test.erl, line 71)
in call from eunit_proc:run_test/1 (eunit_proc.erl, line 510)
in call from eunit_proc:with_timeout/3 (eunit_proc.erl, line 335)
in call from eunit_proc:handle_test/2 (eunit_proc.erl, line 493)
in call from eunit_proc:tests_inorder/3 (eunit_proc.erl, line 435)
in call from eunit_proc:with_timeout/3 (eunit_proc.erl, line 325)
**error:{assertEqual,[{module,mem3_bdu_test},
{line,206},
{expression,"Code"},
{expected,201},
{value,409}]}
output:<<"">>
|
| |
|
|
|
| |
Depending on configuration, it is possible for the shards db to be
different than `_dbs`.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Prevent failures like this:
mem3_sync_event_listener:267: should_set_sync_delay...*failed*
in function gen_server:call/3 (gen_server.erl, line 214)
in call from mem3_sync_event_listener:'-should_set_sync_delay/1-fun-1-'/1
(src/mem3_sync_event_listener.erl, line 268)
**exit:{{noproc,{gen_server,call,
[couch_log_server,
{log,{log_entry,notice,<0.31789.5>,
["config",58,32,91,[...]|...],
"--------",
["2021",45,[...]|...]}}]}},
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Our existing logic for handling rewinds in the changes feed addresses
the following cases:
- A node that contributed to a sequence is in maintenance mode
- A shard that contributed to a sequence has been split
This patch adds support for cases where the node that contributed to a
client-supplied sequence is down at the beginning of the request
handling. It reuses the same logic as the maintenance mode case as these
two situations really ought to be handled the same way.
A future improvement would be to unify the "node down" and "shard split"
logic so that we could handle the compound case, e.g. replacing a shard
from a down node with a pair of shards from nodes that cover the same
range.
Fixes #3788
Co-authored-by: Nick Vatamaniuc <vatamane@gmail.com>
|
| |
|
|
|
|
|
|
|
| |
- Unused functions in `couch_util_tests`
- Unused variables in `couch_prometheus_e2e_tests`
- Unused variable in `dreyfus_blacklist_await_test`
- Deprecated BIF `erlang:now/0` in `dreyfus_purge_test`
- `export_all` flag in dreyfus tests
- Unused variable in `mem3_reshard_test`
|
| |
|
|
|
|
| |
The test only checks that we can update the shard doc so we just verifythat.
Apparently, it doesn't mean we can synchronously access the newly created db
info right away so we just skip that part to avoid a flaky failure.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Noticed this flaky test show up in a few recent test runs, for
[example](https://ci-couchdb.apache.org/blue/organizations/jenkins/jenkins-cm1%2FPullRequests/detail/PR-3799/1/pipeline)
The test was flaky as We were only waiting for the replication task or
scheduler job to appear in the list but didn't not wait until the
value of the task had been updated to an expected value. So the task
might have appeared but then only half the docs written (say, 5
instead of 10). Testing the value at that stage is too early and the
test would fail.
To fix the issue, besides waiting on the task/job to appear in the
list, also wait until its `docs_written` value matches the expected
value. By that point `docs_read` should have caught up as well.
|
| |
|
|
|
|
|
| |
Fetch the libicu base version as well as the collator version. The
base version may be used to determine which libicu library CouchDB is
using. The collator version may be used to debug view behavior in case
when collation order had changed from one version ot the next.
|
| |
|
|
|
|
|
|
|
| |
This fixes issue: https://github.com/apache/couchdb/issues/3786
In addition, add few _all_dbs limit tests since we didn't seem to have
any previously to catch such issues. Plus, test some of the corner
cases which should be caught by the BDU and should return a 403 error
code.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This means `update_seq` values from `GET $db` `last_seq` returned from ` GET
$db/_changes?since=now&limit=` will be more resilient to change feed rewinds.
Besides, those sequences will now be more consistent and users won't have to
wonder why one opaque sequence works slightly differently than another opaque
update sequence.
Previously, when the sequences were returned only as numeric values, it was
impossible to calculate replacements and change feeds had to always rewind back
to 0 for those ranges. With uuids and epochs in play, it is possible to figure
out that some shards might have moved to new nodes or find internal replication
checkpoints to avoid streaming changes feeds from 0 on those ranges.
Some replication Elixir tests decode update sequences, so those were updated to
handle the new uuid and epoch format as well.
Fixes: https://github.com/apache/couchdb/issues/3787
Co-author: Adam Kocoloski kocolosk@apache.org
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
couch_icu_driver is only used for binary strings comparison in
couch_ejson_compare when expression depth becomes greater than 10.
The logic for string comparison is identical to what couch_ejson_compare uses,
so opt to just use couch_ejson_compare instead of keeping a whole other binary
collation driver around.
To avoid a possible infinite loop if couch_ejson_compare nif fails to
load, throw a nif loading error as is common for nif modules.
To avoid another case of a possible infinite retry from of badarg
generated by max depth, and/or an actual bad ejson term, use a
specific max depth error so we don't have to guess when we catch it
and retry term traversal in erlang.
There was another uncodumented case when badarg was thrown besides max
depth or an invalid arg. It was when a prop value was compared with
any other supported type. In erlang it would be handled in these
clauses:
```
less_erl({A},{B}) when is_list(A), is_list(B) -> less_props(A,B);
less_erl({A},_) when is_list(A) -> -1;
less_erl(_,{B}) when is_list(B) -> 1.
```
However, in C we threw a badarg for the last two clauses and relied on
erlang to do all the work. This case was a potential performance issue
as well since that is a common comparison for mango where we may
compare keys against the max json object value (<<255,255,255,255>>).
Add a few property tests in order to validate collation behavior. The two main
ones are:
1) Given an expected sort order of some test values, assert that both the
erlang and nif collators would correctly order any of those test values.
2) In general, the nif collator would sort any json the same way as the
erlang one.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
`infinity` it turns out is a valid configuration value for fabric
request_timeout. We can pass that to Erlang `receive` statement, any arithmetic
with it would fail.
To guard against the crash use the max small int value (60 bits). With enough
shards, due to the exponential nature of the algorithm, we still get a nice
progression from the minimum 100 msec all the way up to the large int value.
This case is illustrated in the test.
Issue: https://github.com/apache/couchdb/issues/3789
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
| |
Prevent this race condition:
*** context setup failed ***
**in function couch_replicator_doc_processor:setup/0 (src/couch_replicator_doc_processor.erl, line 872)
**error:{badmatch,{error,{already_started,<0.4946.0>}}}
|
| |
|
|
|
|
| |
- Prepend unused variable with underscore
- Add nowarn_export_all compiler option
- Use STACKTRACE macro
|
| |
|
|
| |
Delete unused function and remove unused variable.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In the previous attempt [1] we improved the logic by spawning workers on the
matching target shards only. However, that wasn't enough as workers might still
reject the passed in sequence from the old node when it asserts ownership
locally on each shard.
Re-use the already existing replacement clause, where after uuid is matched, we
try harder to find the highest viable sequence. To use the unit test setup as
an example, if the shard moved from node1 to node2, and recorded epoch `{node2,
10}` on the new node, then a sequence generated on node1 before the move, for
example 12, would rewind down to 10 only when calculated on new location on
node2, instead of being rewound all the way down to 0.
[1] https://github.com/apache/couchdb/commit/e83935c7f8c3e47b47f07f22ece327f6529d4da0
|
| |
|
|
| |
To include another changelog entry: https://github.com/apache/couchdb-documentation/commit/4f00da0b0cedf63ebf391e43b1a56bb36f7d0f96
|
| |
|
| |
Missed file in f85cff669f20cee0a54da7bb8c645dfc4d2de5c9
|
| |
|
|
|
|
| |
Based off of the upstream 1.0.9 + CouchDB clone changes
https://github.com/apache/couchdb-jiffy/releases/tag/CouchDB-1.0.9-1
|
| | |
|
| |
|
|
|
|
| |
Fauxton was failing so we backported the fix from main
https://github.com/apache/couchdb/commit/f85cff669f20cee0a54da7bb8c645dfc4d2de5c9
for it.
|
| | |
|
