Backport the new BackgroundPsql modules and the constructor functions,
background_psql() and interactive_psql, to all supported
branches. That makes it easier to backpatch tests that use it.
BackgroundPsql was introduced in version 16. On version 16, this
commit backports just the new timeout argument from master (commit
334f512f45). On older branches, the whole facility. This includes the
change to `use warnings FATAL => 'all'`, which we haven't otherwise
backported, but it seems good to keep the file identical across
branches.
Discussion: https://www.postgresql.org/message-id/b7c64f20-ea01-4f15-9088-0cd6832af149@iki.fi
The same routine to check if a specific pattern can be found in the
server logs was copied over four different test scripts. This refactors
the whole to use a single routine located in PostgreSQL::Test::Cluster,
named log_contains, to grab the contents of the server logs and check
for a specific pattern.
On HEAD, the code previously used assumed that slurp_file() could not
handle an undefined offset, setting it to zero, but slurp_file() does
do an extra fseek() before retrieving the log contents only if an offset
is defined. In two places, the test was retrieving the full log
contents with slurp_file() after calling substr() to apply an offset,
ignoring that slurp_file() would be able to handle that.
Backpatch all the way down to ease the introduction of new tests that
could rely on the new routine.
Author: Vignesh C
Reviewed-by: Andrew Dunstan, Dagfinn Ilmari Mannsåker, Michael Paquier
Discussion: https://postgr.es/m/CALDaNm0YSiLpjCmajwLfidQrFOrLNKPQir7s__PeVvh9U3uoTQ@mail.gmail.com
Backpatch-through: 11
This commit refactors a bit the code in charge of checking for log
patterns when connections fail or succeed, by moving the log pattern
checks into their own routine, for clarity. This has come up as
something to improve while discussing the refactoring of find_in_log().
Backpatch down to 14 where these routines are used, to ease the
introduction of new tests that could rely on them.
Author: Vignesh C, Michael Paquier
Discussion: https://postgr.es/m/CALDaNm0YSiLpjCmajwLfidQrFOrLNKPQir7s__PeVvh9U3uoTQ@mail.gmail.com
Backpatch-through: 14
Currently there is a race condition where if concurrent TAP tests both
test that they can open a port they will assume that it is free and use
it, causing one of them to fail. To prevent this we record a reservation
using an exclusive lock, and any TAP test that discovers a reservation
checks to see if the reserving process is still alive, and looks for
another free port if it is.
Ports are reserved in a directory set by the environment setting
PG_TEST_PORT_DIR, or if that doesn't exist a subdirectory of the top
build directory as set by Makefile.global, or its own
tmp_check directory.
The prove_check recipe in Makefile.global.in is extended to export
top_builddir to the TAP tests. This was already exported by the
prove_installcheck recipes.
Per complaint from Andres Freund
Backpatched from 9b4eafcaf4 to all live branches
Discussion: https://postgr.es/m/20221002164931.d57hlutrcz4d2zi7@awork3.anarazel.de
Remove the need to edit back-branch-specific code sites when
back-patching the addition of a PostgreSQL::Test::Utils symbol. Replace
per-symbol, incomplete alias lists. Give old and new package names the
same EXPORT and EXPORT_OK semantics. Back-patch to v10 (all supported
versions).
Reviewed by Andrew Dunstan.
Discussion: https://postgr.es/m/20220622072144.GD4167527@rfd.leadboat.com
Commit b3b4d8e68a moved our perl test modules to a better namespace
structure, but this has made life hard for people wishing to backpatch
improvements in the TAP tests. Here we alleviate much of that difficulty
by implementing the new module names on top of the old modules, mostly
by using a little perl typeglob aliasing magic, so that we don't have a
dual maintenance burden. This should work both for the case where a new
test is backpatched and the case where a fix to an existing test that
uses the new namespace is backpatched.
Reviewed by Michael Paquier
Per complaint from Andres Freund
Discussion: https://postgr.es/m/20220418141530.nfxtkohefvwnzncl@alap3.anarazel.de
Applied to branches 10 through 14
The previous method for doing that was to write zeroes into a
predetermined set of page locations. However, there's a roughly
1-in-64K chance that the existing checksum will match by chance,
and yesterday several buildfarm animals started to reproducibly
see that, resulting in test failures because no checksum mismatch
was reported.
Since the checksum includes the page LSN, test success depends on
the length of the installation's WAL history, which is affected by
(at least) the initial catalog contents, the set of locales installed
on the system, and the length of the pathname of the test directory.
Sooner or later we were going to hit a chance match, and today is
that day.
Harden these tests by specifically inverting the checksum field and
leaving all else alone, thereby guaranteeing that the checksum is
incorrect.
In passing, fix places that were using seek() to set up for syswrite(),
a combination that the Perl docs very explicitly warn against. We've
probably escaped problems because no regular buffered I/O is done on
these filehandles; but if it ever breaks, we wouldn't deserve or get
much sympathy.
Although we've only seen problems in HEAD, now that we recognize the
environmental dependencies it seems like it might be just a matter
of time until someone manages to hit this in back-branch testing.
Hence, back-patch to v11 where we started doing this kind of test.
Discussion: https://postgr.es/m/3192026.1648185780@sss.pgh.pa.us
Slow hosts may avoid load-induced, spurious failures by setting
environment variable PG_TEST_TIMEOUT_DEFAULT to some number of seconds
greater than 180. Developers may see faster failures by setting that
environment variable to some lesser number of seconds. In tests, write
$PostgreSQL::Test::Utils::timeout_default wherever the convention has
been to write 180. This change raises the default for some briefer
timeouts. Back-patch to v10 (all supported versions).
Discussion: https://postgr.es/m/20220218052842.GA3627003@rfd.leadboat.com
Following migration of Windows buildfarm members running TAP tests to
use of ucrt64 perl for those tests, special processing for msys perl is
no longer necessary and so is removed.
Backpatch to release 10
Discussion: https://postgr.es/m/c65a8781-77ac-ea95-d185-6db291e1baeb@dunslane.net
Commits 6c4a8903b et al. had a couple of deficiencies:
* The logic I added to Cluster::start to see if a PID file is present
could be fooled by a stale PID file left over from a previous
postmaster. To fix, if we're not sure whether we expect to find a
running postmaster or not, validate the PID using "kill 0".
* 017_shm.pl has a loop in which it just issues repeated Cluster::start
calls; this will fail if some invocation fails but leaves self->_pid
set. Per buildfarm results, the above fix is not enough to make this
safe: we might have "validated" a PID for a postmaster that exits
immediately after we look. Hence, match each failed start call with
a stop call that will get us back to the self->_pid == undef state.
Add a fail_ok option to Cluster::stop to make this work.
Discussion: https://postgr.es/m/CA+hUKGKV6fOHvfiPt8=dOKzvswjAyLoFoJF1iQXMNpi7+hD1JQ@mail.gmail.com
"pg_ctl start" might start a new postmaster and then return failure
anyway, for example if PGCTLTIMEOUT is exceeded. If there is a
postmaster there, it's still incumbent on us to shut it down at
script end, so check for the PID file even though we are about
to fail.
This has been broken all along, so back-patch to all supported branches.
Discussion: https://postgr.es/m/647439.1642622744@sss.pgh.pa.us
Commits fdd965d07 and 3cd9c3b92 tested CREATE INDEX CONCURRENTLY by
launching two separate pgbench runs concurrently. This was needed so
that only a single client thread would run CREATE INDEX CONCURRENTLY,
avoiding deadlock between two CICs. However, there's a better way,
which is to use an advisory lock to prevent concurrent CICs. That's
better in part because the test code is shorter and more readable, but
mostly because it automatically scales things to launch an appropriate
number of CICs relative to the number of INSERT transactions.
As committed, typically half to three-quarters of the CIC transactions
were pointless because the INSERT transactions had already stopped.
In passing, remove background_pgbench, which was added to support
these tests and isn't needed anymore. We can always put it back
if we find a use for it later.
Back-patch to v12; older pgbench versions lack the
conditional-execution features needed for this method.
Tom Lane and Andrey Borodin
Discussion: https://postgr.es/m/139687.1635277318@sss.pgh.pa.us
CIC and REINDEX CONCURRENTLY assume backends see their catalog changes
no later than each backend's next transaction start. That failed to
hold when a backend absorbed a relevant invalidation in the middle of
running RelationBuildDesc() on the CIC index. Queries that use the
resulting index can silently fail to find rows. Fix this for future
index builds by making RelationBuildDesc() loop until it finishes
without accepting a relevant invalidation. It may be necessary to
reindex to recover from past occurrences; REINDEX CONCURRENTLY suffices.
Back-patch to 9.6 (all supported versions).
Noah Misch and Andrey Borodin, reviewed (in earlier versions) by Andres
Freund.
Discussion: https://postgr.es/m/20210730022548.GA1940096@gust.leadboat.com
Several TAP tests use poll_query_until() to wait for the postmaster
to restart. They were checking to see if a trivial query
(e.g. "SELECT 1") succeeds. However, that's problematic in the wake
of commit 11e9caff8, because now that we feed said query to psql
via stdin, we risk IPC::Run whining about a SIGPIPE failure if psql
quits before reading the query. Hence, we can't use a nonempty
query in cases where we need to wait for connection failures to
stop happening.
Per the precedent of commits c757a3da0 and 6d41dd045, we can pass
"undef" as the query in such cases to ensure that IPC::Run has
nothing to write. However, then we have to say that the expected
output is empty, and this exposes a deficiency in poll_query_until:
if psql fails altogether and returns empty stdout, poll_query_until
will treat that as a success! That's because, contrary to its
documentation, it makes no actual check for psql failure, looking
neither at the exit status nor at stderr.
To fix that, adjust poll_query_until to insist on empty stderr as
well as a stdout match. (I experimented with checking exit status
instead, but it seems that psql often does exit(1) in cases that we
need to consider successes. That might be something to fix someday,
but it would be a non-back-patchable behavior change.)
Back-patch to v10. The test cases needing this exist only as far
back as v11, but it seems wise to keep poll_query_until's behavior
the same in v10, in case we back-patch another such test case in
future. (9.6 does not currently need this change, because in that
branch poll_query_until can't be told to accept empty stdout as
a success case.)
Per assorted buildfarm failures, mostly on hoverfly.
Discussion: https://postgr.es/m/CAA4eK1+zM6L4QSA1XMvXY_qqWwdUmqkOS1+hWvL8QcYEBGA1Uw@mail.gmail.com
The Msys shell mangles certain patterns in its command line, so avoid
handing arbitrary SQL to psql on the command line and instead use
IPC::Run's redirection facility for stdin. This pattern is already
mostly whats used, but query_poll_until() was not doing the right thing.
Problem discovered on the buildfarm when a new TAP test failed on msys.
Older versions of perl on Windows don't like the list form of pipe open,
and perlcritic doesn't like the string form of open, so we avoid both
with a simpler formulation using qx{}.
Per complaint from Amit Kapila.
Also "make reformat-dat-files".
The only change worthy of note is that pgindent messed up the formatting
of launcher.c's struct LogicalRepWorkerId, which led me to notice that
that struct wasn't used at all anymore, so I just took it out.
A new PostgresVersion object type is created and this is used in
PostgresNode using the output of `pg_config --version` and the result
stored in the PostgresNode object. This object can be compared to other
PostgresVersion objects, or to a number or string.
PostgresNode is currently believed to be compatible with versions down
to release 12, so PostgresNode will issue a warning if used with a
version prior to that.
No attempt has been made to deal with incompatibilities in older
versions - that remains work to be undertaken in a subsequent
development cycle.
Based on code from Mark Dilger and Jehan-Guillaume de Rorthais.
Discussion: https://postgr.es/m/a80421c0-3d7e-def1-bcfe-24777f15e344@dunslane.net
Commit b34ca595ab provided for installation-aware instances of
PostgresNode. However, it turns out that IPC::Run works against this by
caching the path to a binary and not consulting the path again, even if
it has changed. We work around this by calling Postgres binaries with
the installed path rather than just a bare name to be looked up in the
environment path, if there is an installed path. For the common case
where there is no installed path we continue to use the bare command
name.
Diagnosis and solution from Mark Dilger
Discussion: https://postgr.es/m/E8F512F8-B4D6-4514-BA8D-2E671439DA92@enterprisedb.com
In order to avoid getting old logfile contents certain functions in
PostgresNode were doing one of two things. On Windows it rotated the
logfile and restarted the server, while elsewhere it truncated the log
file. Both of these are unnecessary. We borrow from the buildfarm which
does this instead: note the size of the logfile before we start, and
then when fetching the logfile skip to that position before accumulating
contents. This is spelled differently on Windows but the effect is the
same. This is largely centralized in TestLib's slurp_file function,
which has a new optional parameter, the offset to skip to before
starting to reading the file. Code in the client becomes much neater.
Backpatch to all live branches.
Michael Paquier, slightly modified by me.
Discussion: https://postgr.es/m/YHajnhcMAI3++pJL@paquier.xyz
The truncation of the log file, that this set of tests relies on to make
sure that a connection attempt matches with its expected backend log
pattern, fails, as reported by buildfarm member fairywren. Instead of a
truncation, do a rotation of the log file and restart the node. This
will ensure that the connection attempt data is unique for each test.
Discussion: https://postgr.es/m/YG05nCI8x8B+Ad3G@paquier.xyz
The "authenticated identity" is the string used by an authentication
method to identify a particular user. In many common cases, this is the
same as the PostgreSQL username, but for some third-party authentication
methods, the identifier in use may be shortened or otherwise translated
(e.g. through pg_ident user mappings) before the server stores it.
To help administrators see who has actually interacted with the system,
this commit adds the capability to store the original identity when
authentication succeeds within the backend's Port, and generates a log
entry when log_connections is enabled. The log entries generated look
something like this (where a local user named "foouser" is connecting to
the database as the database user called "admin"):
LOG: connection received: host=[local]
LOG: connection authenticated: identity="foouser" method=peer (/data/pg_hba.conf:88)
LOG: connection authorized: user=admin database=postgres application_name=psql
Port->authn_id is set according to the authentication method:
bsd: the PostgreSQL username (aka the local username)
cert: the client's Subject DN
gss: the user principal
ident: the remote username
ldap: the final bind DN
pam: the PostgreSQL username (aka PAM username)
password (and all pw-challenge methods): the PostgreSQL username
peer: the peer's pw_name
radius: the PostgreSQL username (aka the RADIUS username)
sspi: either the down-level (SAM-compatible) logon name, if
compat_realm=1, or the User Principal Name if compat_realm=0
The trust auth method does not set an authenticated identity. Neither
does clientcert=verify-full.
Port->authn_id could be used for other purposes, like a superuser-only
extra column in pg_stat_activity, but this is left as future work.
PostgresNode::connect_{ok,fails}() have been modified to let tests check
the backend log files for required or prohibited patterns, using the
new log_like and log_unlike parameters. This uses a method based on a
truncation of the existing server log file, like issues_sql_like().
Tests are added to the ldap, kerberos, authentication and SSL test
suites.
Author: Jacob Champion
Reviewed-by: Stephen Frost, Magnus Hagander, Tom Lane, Michael Paquier
Discussion: https://postgr.es/m/c55788dd1773c521c862e8e0dddb367df51222be.camel@vmware.com
This type of failure is similar to what has been fixed in c757a3da,
where an authentication failure combined with psql pushing a command
down its communication pipe causes a test failure. This routine is
designed to fail, so sending a query has little sense anyway.
Per buildfarm members gaur and hoverfly, based on an analysis and fix
from Tom Lane.
Discussion: https://postgr.es/m/513200.1617634642@sss.pgh.pa.us
This commit refactors more TAP tests to adapt with the recent
introduction of connect_ok() and connect_fails() in PostgresNode,
introduced by 0d1a3343. This changes the following test suites to use
the same code paths for connection checks:
- Kerberos
- LDAP
- SSL
- Authentication
Those routines are extended to be able to handle optional parameters
that are set depending on each suite's needs, as of:
- custom SQL query.
- expected stderr matching pattern.
- expected stdout matching pattern.
The new design is extensible with more parameters, and there are some
plans for those routines in the future with checks based on the contents
of the backend logs.
Author: Jacob Champion, Michael Paquier
Discussion: https://postgr.es/m/d17b919e27474abfa55d97786cb9cfadfe2b59e9.camel@vmware.com
test_connect_ok() and test_connect_fails() have always been part of the
SSL tests, and check if a connection to the backend should work or not,
and there are sanity checks done on specific error patterns dropped by
libpq if the connection fails.
This was fundamentally wrong on two aspects. First, SSLServer.pm works
mostly on setting up and changing the SSL configuration of a
PostgresNode, and has really nothing to do with the client. Second,
the situation became worse in light of b34ca595, where the SSL tests
would finish by using a psql command that may not come from the same
installation as the node set up.
This commit moves those client routines into PostgresNode, making easier
the refactoring of SSLServer to become more SSL-implementation aware.
This can also be reused by the ldap, kerberos and authentication test
suites for connection checks, and a follow-up patch should extend those
interfaces to match with backend log patterns.
Author: Michael Paquier
Reviewed-by: Andrew Dunstan, Daniel Gustafsson, Álvaro Herrera
Discussion: https://postgr.es/m/YGLKNBf9zyh6+WSt@paquier.xyz
Currently instances of PostgresNode find their Postgres executables in
the PATH of the caller. This modification allows for instances that know
the installation path they are supposed to use, and the module adjusts
the environment of methods that call Postgres executables appropriately.
This facility is activated by passing the installation path to the
constructor:
my $node = PostgresNode->get_new_node('mynode',
installation_path => '/path/to/installation');
This makes a number of things substantially easier, including
. testing third party modules
. testing different versions of postgres together
. testing different builds of postgres together
Discussion: https://postgr.es/m/a94c74f9-6b71-1957-7973-a734ea3cbef1@dunslane.net
Reviewed-By: Alvaro Herrera, Michael Paquier, Dagfinn Ilmari Mannsåker
The existing test script does run pg_basebackup with the -Ft option,
but it makes no real attempt to verify the sanity of the results.
We wouldn't know if the output is incompatible with standard "tar"
programs, nor if the server fails to start from the restored output.
Notably, this means that xlog.c's read_tablespace_map() is not being
meaningfully tested, since that code is used only in the tar-format
case. (We do have reasonable coverage of restoring from plain-format
output, though it's over in src/test/recovery not here.)
Hence, attempt to untar the output and start a server from it,
rather just hoping it's OK.
This test assumes that the local "tar" has the "-C directory"
switch. Although that's not promised by POSIX, my research
suggests that all non-extinct tar implementations have it.
Should the buildfarm's opinion differ, we can complicate the
test a bit to avoid requiring that.
Possibly this should be back-patched, but I'm unsure about
whether it could work on Windows before d66b23b03.
Commit 151c0c5f7 neglected the possibility that a TEMP_CONFIG file
would explicitly set max_wal_senders=0; as indeed buildfarm member
thorntail does, so that it can test wal_level=minimal in other test
suites. Hence, rather than assuming that max_wal_senders=10 will
prevail if we say nothing, set it explicitly.
Set max_replication_slots=10 explicitly too, just to be safe.
Back-patch to v10, like the previous patch.
Discussion: https://postgr.es/m/723911.1601417626@sss.pgh.pa.us
PostgresNode.pm set "max_wal_senders = 5" for replication testing,
but this seems to be slightly too low for our current test suite.
Slower buildfarm members frequently report "number of requested standby
connections exceeds max_wal_senders" failures, due to old walsenders
not exiting instantaneously. Usually, the test does not fail overall
because of automatic walreceiver restart, but sometimes the failure
becomes visible; and in any case such retries slow down the test.
That value came in with commit 89ac7004d, but was soon obsoleted by
f6d6d2920, which raised the built-in default from zero to 10; so that
PostgresNode.pm is actually setting it to less than the conservative
built-in default. That seems pretty pointless, so let's remove the
special setting and let the default prevail, in hopes of making
the TAP tests more robust.
Likewise, the setting "max_replication_slots = 5" is obsolete and
can be removed.
While here, reverse-engineer a comment about why we're choosing
less-than-default values for some other settings.
(Note: before v12, max_wal_senders counted against max_connections
so that the latter setting also needs some fiddling with.)
Back-patch to v10 where the subscription tests were added.
It's likely that the older branches aren't pushing the boundaries
of max_wal_senders, but I'm disinclined to spend time trying to
figure out exactly when it started to be a problem.
Discussion: https://postgr.es/m/723911.1601417626@sss.pgh.pa.us
Reuse cautionary language from src/test/ssl/README in
src/test/kerberos/README. SLRUs have had access to six-character
segments names since commit 73c986adde,
and recovery stopped calling HeapTupleHeaderAdvanceLatestRemovedXid() in
commit 558a9165e0. The other corrections
are more self-evident.
Buildfarm results now imply that Perl's IPC::Run does CRLF conversion
for us if we're using native Perl, but not when using MSys Perl.
Restrict the conversions done by PostgresNode.pm to act only in the
latter case. (Similar conversions done in TestLib.pm and RewindTest.pm
were already handled this way.)
Discussion: https://postgr.es/m/412ae8da-76bb-640f-039a-f3513499e53d@gmx.net
The previous approach was to search-and-destroy all \r occurrences
no matter what. That seems more likely to hide bugs than anything
else; indeed it seems to be hiding one now. Fix things so that
we only transform \r\n to \n.
Side effects: must do this before, not after, chomp'ing if we're
going to chomp, else we'd fail to clean up a trailing \r\n. Also,
remove safe_psql's redundant repetition of what psql already did;
else it might reduce \r\r\n to \n, which is exactly the scenario
I'm hoping to expose.
Perhaps this should be back-patched, but for now I'm content to
see what happens in HEAD.
Discussion: https://postgr.es/m/412ae8da-76bb-640f-039a-f3513499e53d@gmx.net
In commit 850196b610 I (Álvaro) failed to handle the case of walsender
shutting down on an error before setting up its 'xlogreader' pointer;
the error handling code dereferences the pointer, causing a crash.
Fix by testing the pointer before trying to dereference it.
Kyotaro authored the code fix; I adopted Nathan's test case to be used
by the TAP tests and added the necessary PostgresNode change.
Reported-by: Nathan Bossart <bossartn@amazon.com>
Author: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/C04FC24E-903D-4423-B312-6910E4D846E5@amazon.com
Includes some manual cleanup of places that pgindent messed up,
most of which weren't per project style anyway.
Notably, it seems some people didn't absorb the style rules of
commit c9d297751, because there were a bunch of new occurrences
of function calls with a newline just after the left paren, all
with faulty expectations about how the rest of the call would get
indented.
The test suites currently don't use Unix-domain sockets on Windows.
This optionally allows enabling that by setting the environment
variable PG_TEST_USE_UNIX_SOCKETS.
This should currently be considered experimental. In particular,
pg_regress.c contains some comments that the cleanup code for
Unix-domain sockets doesn't work correctly under Windows, which hasn't
been an problem until now. But it's good enough for locally
supervised testing of the functionality.
Reviewed-by: Andrew Dunstan <andrew.dunstan@2ndquadrant.com>
Discussion: https://www.postgresql.org/message-id/flat/54bde68c-d134-4eb8-5bd3-8af33b72a010@2ndquadrant.com
