public.commandprompt.com: Issueshttps://public.commandprompt.com/https://public.commandprompt.com/favicon.ico?16484970162017-08-02T23:44:00Zpublic.commandprompt.com
Redmine postgresql-snap - Task #5340 (New): Change package namehttps://public.commandprompt.com/issues/53402017-08-02T23:44:00ZIvan Lezhnjov
<p>We need to change the name of the snap packages from Postgresql to Postgres.</p> postgresql-snap - Task #5339 (New): Explore tracks featurehttps://public.commandprompt.com/issues/53392017-06-09T03:47:00ZIvan Lezhnjov
<p>While working on <a class="issue tracker-4 status-3 priority-4 priority-default closed" title="Task: Explore app alias functionality (Resolved)" href="https://public.commandprompt.com/issues/5338">#5338</a> (Explore app alias functionality), I was asked in snapcraft forum if we had any plans to support tracks instead of using version number in a package name.</p>
<p>On a quick look, tracks seems like a central feature of snap's package management design and we most likely will have to start using it.</p>
<p>Essentially, it is a way to catalog software using the following scheme track/risk/branch.</p>
<p>E.g. 9.3/stable/psql-freeze-fix</p>
<p>I will make some tests to get a good idea about how it works, how we could transition from the current approach and start using tracks.</p>
<p>Ivan</p> postgresql-snap - Feature #5327 (In Progress): New snap packages to be createdhttps://public.commandprompt.com/issues/53272016-12-09T07:11:35ZJoshua Drake
<p>pgbouncer<br />pgpool</p> postgresql-snap - Task #5321 (New): Make postgres a systemd daemonhttps://public.commandprompt.com/issues/53212016-10-28T06:34:00ZIvan Lezhnjov
<p>Running postgres process as a daemon via an automatically generated systemd service file, results in snap utility failing to install a snap package.</p>
<p>The reason for this is because the said snap utility attempts to start all daemons when it is installing a snap package. The package is always installed as root and thus any daemon is run as root too. However, PostgreSQL is not designed to run as root and it actually refuses to. Thus, if a daemon startup fails, installation of a snap package fails too.</p>
<p>Ideally, postgres should be a systemd-managed process but given the clash of design philosophies it currently doesn't work well.</p>
<p>We need to start a dialog with snapcraft developers and see if they can propose a solution to this problem that doesn't involve patching PostgreSQL.</p> pitrtools - Feature #5306 (Feedback): Reimplementation of queuehttps://public.commandprompt.com/issues/53062014-05-13T12:58:27ZJoshua A. Drakezam@commandprompt.com
<p>Reimplementation of PITRTools queue<br />The current queue system possesses the following limitations:<br />1. No job control - If we are generating xlogs faster than we can ship them, rsync processes will be spawned uncontrollably.<br />2. A continuous stream of transaction logs must be generated for correct operation of the archiver. If there are no new logs <br />coming in, logs still in the queues will not get shipped.</p>
<p>Suggested solution:<br /><pre>
Queue management daemon - Have cmd_archiver only queue logs, with a separate daemon shipping them.
* cmd_queue daemon
- write PID to a file
- In worker thread
] Ship logs in queue with subprocess calls when self to-do list is not empty
} If successful, remove paths from self to-do list and log appropriate output
} Else, log which queues failed
] Wait a few seconds [configurable]
] Repeat
- In main thread
] Wait for data [absolute paths to queued logs] on a pipe or UNIX socket
] On data, parse and extract paths, append paths to self to-do list
] Repeat
* When cmd_archiver is called with a new archive, copy the archive to slave queues and write paths to cmd_queue via pipe/socket
</pre></p> pitrtools - Discussion #5263 (New): PITRTools ini configuration files are too complexhttps://public.commandprompt.com/issues/52632012-09-17T04:03:52ZAlexey Klyukinalexk@commandprompt.com
<p>Since we position PITRTools as a tool to simplify WAL-shipping and replication setup, I though I'd count the number of options once needs to set in PIRTools, against pristine PostgresSQL:</p>
<p>PITRTools:<br />Archiver: 15<br />Standby: 31.</p>
<p>Comparing to the pristine PostgreSQL:<br />primary: (wal_level, archive_mode, archive_command, max_wal_senders) - 4.</p>
<p>standby (most of them belong to recovery.conf): (hot_standby, standby_mode, primary_conninfo, restore_command, archive_cleanup_command): 5</p>
<p>We need to rethink PITRTools configuration options. I've made the first step by splitting them into required and uncommon, and put reasonable defaults for the second ones (i.e. assumed ssh and rsync path is fixed, and there is no need to tweak rsync options or set ssh_debug), but to make the tool easier than pristine PostgreSQL we need to put some code to autodetect some of the parameters (i.e. autodetect things like PostgreSQL version, checkpoint_segments value for 8.2 and below, rsync_version, etc.)</p> PL/php - Discussion #5256 (New): Creating a PL/php distributionhttps://public.commandprompt.com/issues/52562011-05-21T18:36:16ZJoshua Drake
<p>Since we need PHP source to create plphp (due to no one shipping as embed), perhaps we should just include the source as part of the download of php and have a predefined configure that deals with all of it for people. Thoughts?</p> Pggraph - Feature #4915 (In Progress): insert_time should be timestamptzhttps://public.commandprompt.com/issues/49152007-03-21T06:35:32ZDarcy Buskermolendarcyb@foo.bar
<p>We should be providing all date logging with timestamps, so that it's easier to work in our clients respective timezones</p> Pggraph - Feature #4914 (In Progress): customer feedbackhttps://public.commandprompt.com/issues/49142007-03-19T20:54:29Zbford -bford@foo.bar
<p>(08:53:42 PM) linuxpoet: oh you mean like a custom repot?<br />(08:53:51 PM) Crad: Something like <a class="external" href="http://server12.uac.ehpg.net/rtg/view.php?rid=1">http://server12.uac.ehpg.net/rtg/view.php?rid=1</a> would be awesome<br />(08:54:00 PM) Crad: even if I had to pick the tables to show.<br />(08:54:34 PM) linuxpoet: yeah we could probably do something like that<br />(08:54:41 PM) linuxpoet: let me make a ticket so we don't loose it</p> Pggraph - Feature #4908 (In Progress): Table statistics tablenamehttps://public.commandprompt.com/issues/49082007-01-07T19:26:28Zbford -bford@foo.bar
<p>We need to consider drawing this during initial load or reworking the menu set so that the user selects their relation which then loads the table name that they can graph off if.</p> Pggraph - Feature #4890 (In Progress): new report(s)https://public.commandprompt.com/issues/48902006-09-25T22:36:40ZDarcy Buskermolendarcyb@foo.bar
<p>Create a new report that compares the relation size to the number of tuples in the relation, this can be used to help show relation bloat over time. This report(s) should contain indexes as well as tables.</p> PL/php - Feature #4984 (In Progress): Composite Type Argument as tablehttps://public.commandprompt.com/issues/49842005-12-15T03:38:44Zbford -bford@foo.bar
<p>creating composite type argument named table:</p>
<pre> CREATE TABLE employee (name text, basesalary integer, bonus integer); </pre>
<pre> CREATE TABLE employee_ext (name text, basesalary integer, bonus integer, other text); </pre><br /><pre>
CREATE FUNCTION empcomp(table) RETURNS integer AS $$
return $argsr0['basesalary'] + $argsr0['bonus'];
$$ LANGUAGE 'plphp';
</pre><br />SELECT name, empcomp(table) FROM employee;<br />SELECT name, empcomp(table) FROM employee_ext;
<p>eg. type table is table taken from FROM (so once time will be employee, second time employee_ext)<br />now i must have x functions to do same thing with different argument</p> PL/php - Feature #4979 (In Progress): Add support for IN/OUT parametershttps://public.commandprompt.com/issues/49792005-12-05T07:37:14ZÁlvaro Herreraalvherre@commandprompt.com
<p>Add support for OUT parameters so that the user can skip specifying a return type explicitely.</p> PL/php - Bug #4972 (In Progress): array PHP <-> Pg conversion is brokenhttps://public.commandprompt.com/issues/49722005-11-23T11:48:44ZÁlvaro Herreraalvherre@commandprompt.com
<p>Element parsing in arrays can easily be confused. Test case:</p>
<pre>
create or replace function another_text_test(text[], int)
returns text language plphp as $$ return $argsr0[$argsr1]; $$;
</pre>
<p>Strings should be quoted before passing them to PHP, so the NOTICE from PHP here is bogus:</p>
<pre>
pl_regression=# select another_text_test($${"foo",'bar', "{"}$$, 0);
NOTICE: plphp: Use of undefined constant foo - assumed 'foo'
another_text_test
-------------------
foo
(1 fila)
</pre>
<p>The same reason causes this to "eat" the inner level of quotes:</p>
<pre>
pl_regression=# select another_text_test($${"'bar'", "{"}$$, 0);
another_text_test
-------------------
bar
(1 fila)
</pre>
<p>Note particularly this totally unexpected behavior:</p>
<pre>
pl_regression=# select another_text_test($${"'bar'", "{"}$$, 1);
another_text_test
-------------------
array(
(1 fila)
</pre> PL/php - Feature #4967 (In Progress): Redesign memory handlinghttps://public.commandprompt.com/issues/49672005-11-16T17:31:06ZÁlvaro Herreraalvherre@commandprompt.com
<p>Currently memory is not handled very well. It would be nice if we had one MemoryContext for each function that could be automatically reset or deleted when the function is recompiled. Additionally we would need a compilation context, and maybe an execution context as well.</p>