| Commit message (Collapse) | Author | Age |
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
slf4j and logback were already used in ExoneraTor. This commit makes all
classes use slf4j and replaced j.u.l and System.{err,out} logging.
|
| |
|
|
| |
Also added test sources to checkstyle target.
|
| |
|
|
|
| |
This includes making utils methods static and also moving all
Gson/JSON related code into QueryResponse for better encapsulation.
|
| | |
|
| |
|
|
|
|
| |
We don't expect results to be in order, and taking out ORDER BY
statements means we don't need to worry about replacing numbers with
names for better readability.
|
| | |
|
| | |
|
| |
|
|
| |
Implements #16596.
|
| |
|
|
|
|
|
|
| |
This prepares splitting up ExoneraTor into front-end and back-end.
Note that some code duplication between ExoneraTorServlet and
QueryServlet was deemed acceptable, because ExoneraTorServlet will be
moved to metrics-web in the medium term anyway.
|
| |
|
|
|
|
| |
This is another prerequisite for changing queries towards making a
single query per request (#16596) and towards splitting up ExoneraTor
into front-end and back-end.
|
| |
|
|
|
|
|
|
|
| |
1: Parse the request.
2: Query the database.
3: Write the response.
This doesn't just clean up code, it's also a prerequisite for changing
queries towards making a single query per request (#16596).
|
| |
|
|
| |
Implements #22811.
|
| | |
|
| |
|
|
| |
Implements #22682.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
Language names and translations of "This page is also available in the
following languages:" taken from https://www.debian.org/.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
| |
Apparently, the zero-width space character causes trouble when copying
and pasting fingerprints. Fortunately, we can let the fingerprint
break anywhere using CSS.
Fixes #17231.
|
| | |
|
| |
|
|
| |
Implements #21535.
|
| | |
|
| | |
|
| |
|
|
| |
Removed obsolete git-submodule.
|
| | |
|
| |
|
|
|
|
| |
We're using metrics-lib for downloading descriptors from CollecTor,
but we're still using our own parsing code. Let's avoid duplicating
code by using what metrics-lib provides.
|
| |
|
|
|
|
|
|
| |
Upgrade to metrics-lib 1.4.0 and include newly required logging .jar
files.
Include PostgreSQL .jar file in the .war file rather than relying on
Tomcat to provide it.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Turns out that resource bundles must be ISO 8859-1 encoded, so we'll
have to convert the German file using:
native2ascii -encoding UTF-8 old new
Also make some minor changes to make UTF-8 encoding of the resulting
web page more explicit, and correct the included language code.
Fixes #19954.
|
| |
|
|
|
|
| |
Fix typo "ant" to "and" and add to the end of all sentences.
Both tweaks suggested by John Lorenz.
|
| |
|
|
|
|
|
|
| |
Required changes from previous schema:
- CREATE OR REPLACE FUNCTION insert_statusentry [...];
- CREATE OR REPLACE FUNCTION insert_exitlistentry [...];
- ALTER TABLE statusentry DROP COLUMN descriptor;
- ALTER TABLE exitlistentry DROP COLUMN rawexitlistentry;
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
Copied over from CollecTor.
|
| | |
|
| |
|
|
| |
Implements #19404.
|
| |
|
|
| |
Fixes #18281.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We're using PostgreSQL 9.1 which, apparently, optimizes functions only
once rather than for each call.
From the 9.2 release notes:
"""
Allow the planner to generate custom plans for specific parameter values
even when using prepared statements (Tom Lane)
In the past, a prepared statement always had a single "generic" plan that
was used for all parameter values, which was frequently much inferior to
the plans used for non-prepared statements containing explicit constant
values. Now, the planner attempts to generate custom plans for specific
parameter values. A generic plan will only be used after custom plans have
repeatedly proven to provide no benefit. This change should eliminate the
performance penalties formerly seen from use of prepared statements
(including non-dynamic statements in PL/pgSQL).
"""
This is rather ugly, so maybe we can take this out when upgrading to 9.2
or higher.
Still related to #17488.
|
| |
|
|
|
|
|
|
|
|
|
| |
Queries that match one or more exit list entries took almost forever,
because avoided all the fine indexes we created for it. The workaround is
to search by /24 of the requested IP address (or /48 in case of IPv6)
which we already have in the database. We should want to do this anyway
to avoid making subsequent queries if there's no hit for the exact IP
address that was requested.
Fixes #17488. Thanks, starlight, for all the feedback!
|