Next parallelism improvements will allow pgloader to use more than one
COPY thread to load data, with the impact of changing the order of rows
in the database.
Rather than doing a copy out and `diff` of the data just loaded, load
the reference data and do the diff in SQL:
select * from loaded.data
except
select * from expected.data
If such a query returns any row, we know we didn't load what was
expected and the regression test is failing.
This regression testing facility should also allow us to finally add
support for multiple-table regression tests (sqlite, mysql, etc).
Our PostgreSQL driver uses CFFI to load the SSL support from open ssl
and as a result the certificate and key file names should be strings
rather than pathnames. Should fix#308 again...
This fixes#308 by automatically using the PostgreSQL Client Side SSL
files as documented in the following reference:
http://www.postgresql.org/docs/current/static/libpq-ssl.html#LIBPQ-SSL-FILE-USAGE
This uses the Postmodern special support for it. Unfortunately couldn't
test it locally other than it doesn't break non-ssl connections. Pushing
to have user feedback.
We used to have a reader and a writer cooperating concurrently into
loading the data from the source to PostgreSQL. The tranformation of the
data was then the responsibility of the reader thread.
Measurements showed that the PostgreSQL processes were mostly idle,
waiting for the reader to produce data fast enough.
In this patch we introduce a third worker thread that is responsible for
processing the raw data into pre-formatted batches, allowing the reader
to focus on extracting the data only. We now have two lparallel queues
involved in the processing, the raw queue contains the vectors of raw
data directly, and the processed-queue contains batches of properly
encoded strings for the COPY text protocol.
On the test laptop the performance gain isn't noticeable yet, it might
be that we need much larger data sets to see a gain here. At least the
setup isn't detrimental to performances on smaller data sets.
Next improvements are going to allow more features: specialized batch
retry thread and parallel table copy scheduling for database sources.
Let's also continue caring about performances and play with having
several worker and writer threads for each reader. In later patches.
And some day, too, we will need to make the number of workers a user
defined variable rather than something hard coded as today. It's on the
todo list, meanwhile, dear user, consider changing the (make-kernel 6)
into (make-kernel 12) or something else in src/sources/mysql/mysql.lisp,
and consider enlighting me with whatever it is you find by doing so!
Have each thread publish its own start-time so that the main thread may
compute time spent in source and target processing, in order to fix the
crude hack of taking (max read-time write-time) in the total time column
of the summary.
We still have some strange artefacts here: we consider that the full
processing time is bound to the writer thread (:target), because it
needs to have the reader done already to be able to COPY the last
batch... but in testing I've seen some :source timings higher than the
:target ones...
Let's solve problems one at a time tho, I guess multi-threading and
accurate wall clock times aren't to be expected to mix and match that
easily anyway (multi cores, single RTC and all that).
Now that we can setup many concurrent threads working against the
PostgreSQL database, and before we open the number of workers to our
users, install an heuristic to manage the PostgreSQL error classes “too
many connection” and “configuration limit exceeded” so that pgloader
waits for some time (*retry-connect-delay*) then tries connecting again.
It's quite simplistic but should cover lots of border-line cases way
more nicely than just throwing the interactive debugger at the end user.
The newly added statistics are showing that read+write times are not
enough to explain how long we wait for the data copying, so it must be
the workers setup rather than the workers themselves.
From there, let lparallel work its magic in scheduling the work we do in
parallel in pgloader: rather than doing blocking receive-result calls
for each table, only receive-result at the end of the whole
copy-database processing.
On test data here on the laptop we go from 6s to 3s to migrate the
sakila database from MySQL to PostgreSQL: that's because we have lots of
very small tables, so the cost of waiting after each COPY added up quite
quickly.
In passing, stop sharing the same connection object in between parallel
workers that used to be controlled active in-sequence, see the new API
clone-connection (which takes over new-pgsql-connection).
Add metrics to devise where the time is spent in current pgloader code
so that it's possible to then optimize away the batch processing as we
do it today.
Given the following extract of the measures, it seems that doing the
data transformations in the reader thread isn't so bright an idea. More
to come.
table name total time read write
----------------- -------------- --------- ---------
extract 2.014s
before load 0.050s
fetch 0.000s
----------------- -------------- --------- ---------
geolite.location 16.090s 15.933s 5.732s
geolite.blocks 28.896s 28.795s 5.312s
----------------- -------------- --------- ---------
after load 37.772s
----------------- -------------- --------- ---------
Total import time 1m25.082s 44.728s 11.044s
In order to later be able to have more worker threads sharing the
load (multiple readers and/or writers, maybe more specialized threads
too), have all the stats be managed centrally by a single thread. We
already have a "monitor" thread that get passed log messages so that the
output buffer is not subject to race conditions, extend its use to also
deal with statistics messages.
In the current code, we send a message each time we read a row. In some
future commits we should probably reduce the messaging here to something
like one message per batch in the common case.
Also, as a nice side effect of the code simplification and refactoring
this fixes#283 wherein the before/after sections of individual CSV
files within an ARCHIVE command where not counted in the reporting.
When the list of columns of the PostgreSQL target table isn't given in
the load command, pgloader will happily query the system catalogs to get
that information. The list-columns query didn't get the memo about the
qualified table name format and the with-schema macro... fix#288.
When parsing table names in the target URI, we are careful of splitting
the table and schema name and store them into a cons in that case. Not
all sources methods got the memo, clean that up.
See #182 and #186, a pull request I am now going to be able to accept.
Also see #287 that should be helped by being able to apply #186.
The dry run option will currently only check database connections, but
as that happens after having correctly parsed the load file, it allows
to also check that the command file is correct for the parser.
Note that the list load-data API isn't subject to the dry-run method.
In passing, we add some more API entry points to the connection objects
and we should actually clean the code base to use the new QUERY generic
all over the place. It's for another patch tho.
The COPY TEXT format accepts non printable characters with an escaped
sequence wherin pgloader can pass in the octal number for the character
in its encoding. When doing that with small numbers like \6 and the
non-printable character is then followed by other numbers, then it
becomes e.g. \646 which might not be part of the target encoding...
To fix, always left pad the character octal number with zeroes, so that
we now send in \00646 which COPY knows how to read: the char at \006
then 4 then 6.
Also copy the test case over to pgloader and run it in the test suite.
The user experience is greatly enhanced by this little change, where you
know from the logs that pgloader could actually connect rather than
thinking it might be still trying...
The index and constraint names given by PostgreSQL catalogs should not
be second guessed, we need to just quote them. The identifier down
casing is interesting when we get identifiers from other system for a
migration, but are wrong when dropping existing indexes in PostgreSQL.
Also, the :null special value from Postmodern was routing the code
badly, just transform it manually to nil when fetching the index list,
manually.
MySQL names its primary keys "PRIMARY" and we need to always uniquify
this name even when the used asked pgloader to preserve index names.
Also, the create-indexes-again function now needs to ask for index names
to be preserved specifically.
The new option 'drop indexes' reuses the existing code to build all the
indexes in parallel but failed to properly account for that fact in the
summary report with timings.
While fixing this, also fix the SQL used to re-establish the indexes and
associated constraints to allow for parallel execution, the ALTER TABLE
statements would block in ACCESS EXCLUSIVE MODE otherwise and make our
efforts vain.
PostgreSQL rightfully forbifs DROP INDEX when the index is used to
enforce a constraint, the proper SQL to issue is then ALTER TABLE DROP
CONSTRAINT. Also, in such a case pg_dump issues a single ALTER TABLE ADD
CONSTRAINT statement to restore the situation.
Have pgloader do the same with indexes that are used to back a constraint.
When loading against a table that already has index definitions, the
load can be quite slow. Previous commit introduced a warning in such a
case. This commit introduces the option "drop indexes" that is not used
by default.
When this option is used, pgloader drops the indexes before loading the
data then create the indexes again with the same definitions as before.
All the indexes are created again in parallel to optimize performances.
Only primary key indexes can't be created in parallel, so those are
created in two steps (create unique index then alter table).
Pre-existing indexes will reduce data loading performances and it's
generally better to DROP the index prior to the load and CREATE them
again once the load is done. See #251 for an example of that.
In that patch we just add a WARNING against the situation, the next
patch will also add support for a new WITH clause option allowing to
have pgloader take care of the DROP/CREATE dance around the data
loading.
In some cases (such as when using a very old PostgreSQL instance or an
Amazon Redshift service, as in #255), the function pg_get_keywords()
does not exists but we assume that pgloader might still be able to
complete its job.
We're better off with a static list of keywords than with a unhandled
error here, so let's see what happens next with Redshift.
When given a file in the COPY format, we should expect that its content
is already properly escaped as expected by PostgreSQL. Rather than
unescape the data then escape it again, add a new more of operation to
format-vector-row in which it won't even try to reformat the data.
In passing, fix an off-by-one bug in dealing with non-ascii characters.
We used to parse qualified table names as a simple string, which then
breaks attempts to be smart about how to quote idenfifiers. Some sources
are known to accept dots in quoted table names and we need to be able to
process that properly without tripping on qualified table names too
late.
Current code might not be the best approach as it's just using either a
cons or a string for table names internally, rather than defining a
proper data structure with a schema and a name slot.
Well, that's for a later cleanup patch, I happen to be lazy tonight.
See test/parse/hans.goeuro.load for an example usage of the new option.
In passing, any error when creating indexes is now properly reported and
logged, which was missing previously. Oops.
This option is dangerous and allows to skip ALL triggers when loading
data against PostgreSQL. This includes foreign key constraints
definitions and will allow loading data out of order.
When using both the options "create no table" and "disable triggers" it
will be possible to load data into a schema prepared by your favorite
external tool, at the cost of not validating FK constraints. Use with
care.
Fix#167.
PostgreSQL requires that an idenfitier begin with letters or underscore
only, so an identifier that begins with a digit must be quoted. In the
current coding pgloader will unecessarily quote some identifiers that
begin with a unicode accentuated letter, but that's only cosmetic and
isn't worth worrying about (famous last words).
PostgreSQL COPY format is not really CSV but something way easier to
parse. Funnily enough, parsing it as CSV is not that easy, so we add
here a special simple parser for the COPY format.
It should be quite useful too try loading again reject data files from
pgloader after manual fixing, too. It's still missing some documentation
without any good excuse for that, will add soon.
That's the big refactoring patch I've been sitting on for too long.
First, refactor connection handling to use a uniformed "connection"
concept (class and generic functions API) everywhere, so that the COPY
derived objects just use that in their :source-db and :target-db slots.
Given that, we don't need no messing around with *pgconn* and *myconn-*
and other special variables at all anywhere in the tree.
Second, clean up some oddities accumulated over time, where some parts
of the code didn't get the memo when new API got into place.
Third, fix any other oddity or missing part found while doing those
first two activities, it was long overdue anyway...
In passing, refactor the *pgconn- dynamic bindings in favor of directly
using the connection property list straight from the connection string
parser, processing it when necessary. That allows to make it simple to
add an internal :use-ssl property.
First, the index names in MS SQL, as in MySQL, are only unique per
table, whereas they need to be globally unique (per database) in
PostgreSQL. So reuse the infrastructure we had for MySQL here.
Second, the way we trick table names in index and fkey structures means
that we already did quote the names and we don't want to quote them
again, so add a new possible *identifier-case* value to handle the case
where nothing is to be done, pretty please.
Rather than doing ALTER TABLE directly, use CREATE UNIQUE INDEX in the
all in parallel concurrent index build per table, and only in the end
game "upgrade" that unique index into a PRIMARY KEY using ALTER TABLE.
The reason why it's a good idea to do that is to avoid an ACCESS
EXCLUSIVE LOCK at ALTER TABLE time, which is killing our index build
concurrency.
