These warnings were only being reported to PostgreSQL on the console. Now they are also recorded in the async log increasing the chance that they will be seen.
This also improves coverage by requiring a warning during async processing to have a test case, which has been added.
The tests were originally written by loading values directly into the configuration before the parser was available.
Update to use harnessCfgLoad() to simplify the tests and make them compatible with upcoming config changes.
Return a path missing error when a stanza is specified for the info command but the stanza does not exist in the repository.
Previously [] was returned, which is still the case if no stanza is specified and the repository does not exist.
lstRemoveIdx(list, 0) resulted in the entire list being moved down to the first position which could take a long time for big lists. This is a common pattern in backup/restore when processing file queues.
Instead simply move the list pointer up when first item is removed. Then on insert check if there is space at the beginning when there is no longer space at the end and do the move then. This way if a list is built and then drained without any new inserts then no move is required.
There were a number of places in the code where "hostId" was used, but hostId is just the option group index + 1 so this led to a lot of +1 and -1 to convert the id to an index and vice versa.
Instead just use the zero based index wherever possible. This is pretty much everywhere except when the host-id option is read or set, or where a message is being formatted for the user.
Also fix a bug in protocolRemoteParam() where remotes spawned from the main process could get process ids that were not 0. Only the locals should spawn remotes with process id > 0. This seems to have been harmless since the process id is only a label, but it could be confusing when debugging.
iniLoad() was trimming lines which meant that a leading space would not pass checksum validation when a manifest was reloaded. Remove the trims since files we write should never contain extraneous spaces. This further diverges the format for the functions that read conf files (e.g. pgbackrest.conf) and those that read info (e.g. manifest) files.
While we are at it also allow [ and # as initial characters. # was reserved for comments but we never put comments into info files. [ denotes a section but we can get around this by never allowing arrays as values in info files, so if a line ends in ] it must be a section. This is currently the case but enforce it by adding an assert to info/info.c.
The tests were originally written by loading values directly into the configuration before the parser was available.
Update to use harnessCfgLoadRaw() to simplify the tests and make them compatible with upcoming config changes.
Note that some unreachable conditions were removed since they could not be reached via a parsed config, only by munging values directly into the config. cfgOptionTest(optionId) was removed because a non-default value must always be set. cfgOptionValid(cfgOptLogTimestamp) was removed because it is true for all commands except for cfgCmdNone, which is checked with an assert.
cfgOptionId() did not recognize deprecated options which made the help command throw errors when they were specified on the command line. cfgParseOption() will correctly identify deprecated options.
cfgParseOption() can also be used in cfgParse() to reduce code duplication when parsing info out of the option value returned by optionFind().
Finally, code the option key index separately in parse.auto.c. For now they are simply added back together but future code will need them separated.
This has always been equivalent to the ConfigCommand enum so it just adds complexity.
It was created for symmetry with ConfigDefineOption, which will also be removed soon.
Currently indexes above 1 do not have dependencies checked, so this doesn't error.
In a future commit we will enable those checks and this will error if it is not fixed.
These constants don't scale well as the index total is increased for an option.
The core code rarely uses these options and they are easily replaced with cfgOptionName().
The tests had started to make use of the constants, so provide functions that build the option name from the optionId and, optionally, the optionKey.
WAL timeline history files were not being expired because they were small and generally not very plentiful.
However, in some cases large numbers of history files may be generated so it makes sense to remove useless history files to keep things tidy.
The history file for the oldest retained timeline is kept for debugging purposes even though it is not used for recovery.
Group related options together so operations (e.g. valid, test, index total) can be performed on all options in the group.
Previously, options at the top of the hierarchy of the related options were used to do these tests. This was prone to error as option relationships changed and it was not always clear which option (or options) should be used.
Bug Fixes:
* Error with hints when backup user cannot read pg_settings. (Reviewed by Stefan Fercot, Cynthia Shang. Reported by Mohamed Insaf K.)
Features:
* PostgreSQL 13 support. (Reviewed by Cynthia Shang.)
Improvements:
* Improve PostgreSQL version identification. (Reviewed by Cynthia Shang, Stephen Frost.)
* Improve working directory error message. (Reviewed by Stefan Fercot.)
* Add hint about starting the stanza when WAL segment not found. (Contributed by David Christensen. Reviewed by David Steele.)
* Add hint for protocol version mismatch. (Reviewed by Cynthia Shang. Suggested by loop-evgeny.)
Documentation Improvements:
* Add note that pgBackRest versions must match when running remotely. (Reviewed by Cynthia Shang. Suggested by loop-evgeny.)
* Move info command text to the reference and link to user guide. (Reviewed by Cynthia Shang. Suggested by Christophe Courtois.)
* Update yum repository path for CentOS/RHEL user guide. (Contributed by Heath Lord. Reviewed by David Steele.)
Update the documentation to explicitly state that versions must match across hosts when running remotely.
Add a hint to the protocol version mismatch error to help the user identify the problem.
Add older PostgreSQL versions to the u18 container that were not available before.
This also updates all minor versions for prior versions of PostgreSQL.
Scan the WAL archive for missing or invalid files and build up ranges of WAL that will be used to verify backup integrity. A number of errors and warnings are currently emitted but they should not be considered authoritative (yet).
The command is incomplete so is marked internal.
Previously, catalog versions were fixed for all versions which made maintaining the catalog versions during PostgreSQL beta and release candidate cycles very painful. A version of pgBackRest which was functionally compatible was rendered useless by a catalog version bump in PostgreSQL.
Instead use only the control version to identify a PostgreSQL version when possible. Some older versions require a catalog version to positively identify a PostgreSQL version, so include them when required.
Since the catalog number is required to work with tablespaces it will need to be stored. There's already a copy of it in backup.info so use that (even though we have been ignoring it in the C versions).
These values are not used by the Perl integration tests so maybe it would be better to remove them, but for now just update since they should not be changing again for PG13.
This condition used to give a not-very-clear error which we have been intending to improve. But in the meantime the changes in fbff299 resulted in a segfault for this condition instead because the data_directory was assumed to be non-NULL.
Fix this by explicitly throwing an error with hints when any row in pg_settings cannot be selected.
This file is created by pg_basebackup so might be in the data directory if the cluster was restored from a pg_basebackup backup. Also exclude backup_manifest.tmp since it is possible to find that in the backup directory.
Improve the wording of the error message and add a hint to make it clearer what is wrong and how the user can fix it.
Also change the assert to a regular error since this is not an internal error.
If a stop command has been issued the check command fails due to archiving timing out.
Provide a hint to document this situation and point the user in the proper direction.
If the callback never returned any jobs then protocolParallelDone() would never be true. The reason is that the done state was being set in protocolParallelResult(), which never gets called if there are no results.
Calling protocolParallelResult() doesn't make much sense in this case so instead move the done logic to protocolParallelDone().
For current usage of ProtocolParallel we ensure there are jobs before processing so this is not a live issue, but the new behavior is required for future development.
Bug Fixes:
* Suppress errors when closing local/remote processes. Since the command has completed it is counterproductive to throw an error but still warn to indicate that something unusual happened. (Reviewed by Cynthia Shang. Reported by argdenis.)
* Fix issue with = character in file or database names. (Reviewed by Bastian Wegge, Cynthia Shang. Reported by Brad Nicholson, Bastian Wegge.)
Features:
* Automatically retrieve temporary S3 credentials on AWS instances. (Contributed by David Steele, Stephen Frost. Reviewed by Cynthia Shang, David Youatt, Aleš Zelený, Jeanette Bromage.)
* Add archive-mode option to disable archiving on restore. (Reviewed by Stephen Frost. Suggested by Stephen Frost.)
Improvements:
* PostgreSQL 13 beta3 support. Changes to the control/catalog/WAL versions in subsequent betas may break compatibility but pgBackRest will be updated with each release to keep pace.
* Asynchronous list/remove for S3/Azure storage. (Reviewed by Cynthia Shang, Stephen Frost.)
* Improve memory usage of unlogged relation detection in manifest build. (Reviewed by Cynthia Shang, Stephen Frost, Brad Nicholson, Oscar. Suggested by Oscar, Brad Nicholson.)
* Proactively close file descriptors after forking async process. (Reviewed by Stephen Frost, Cynthia Shang.)
* Delay backup remote connection close until after archive check. (Contributed by Floris van Nee. Reviewed by David Steele.)
* Improve detailed error output. (Reviewed by Cynthia Shang.)
* Improve TLS error reporting. (Reviewed by Cynthia Shang, Stephen Frost.)
Documentation Bug Fixes:
* Add none to compress-type option reference and fix example. (Reported by Ugo Bellavance, Don Seiler.)
* Add missing azure type in repo-type option reference. (Fixed by Don Seiler. Reviewed by David Steele.)
* Fix typo in repo-cipher-type option reference. (Fixed by Don Seiler. Reviewed by David Steele.)
Documentation Improvements:
* Clarify that expire must be run regularly when expire-auto is disabled. (Reviewed by Douglas J Hunley. Suggested by Douglas J Hunley.)
When restoring a cluster that will be promoted but is not intended to be the new primary, it is important to disable archiving to avoid polluting the repository with useless WAL. This option makes disabling archiving a bit easier.
Automatically retrieve the role and temporary credentials for S3 when the AWS instance is associated with an IAM role. Credentials are automatically updated when they are <= 5 minutes from expiring.
Basic configuration is to set repo1-s3-key-type=auto. repo1-s3-role can be used to set a specific role, otherwise it will be retrieved automatically.
Add more info (command, version, options) to asserts, and errors when debug logging is enabled. This won't cover all cases but might mean we get more info in some circumstances.
When testing the common/stack-trace module it is important not to call this test function since the trace stack is empty and it will cause a buffer under run.
Instead use a macro that is only defined under the correct circumstances and add an assert() to catch future regressions.
Currently each module that needs to collect statistics implements custom code to do so. This is cumbersome.
Create a general purpose module for collecting and reporting statistics. Statistics are output in the log at detail level, but there are other uses they could be put to eventually.
No new functionality is added. This is just a drop-in replacement for the current statistics, with the advantage of being more flexible.
The new stats are slower because they involve a list lookup, but performance testing shows stats can be updated at about 40,000/ms which seems fast enough for our purposes.
HTTP/1.0 connections are closed by default after a single response. Other than that, treat 1.0 the same as 1.1.
HTTP/1.0 allows different date formats that we can't parse but for now, at least, we don't need any date headers from 1.0 requests.
The prior implementation only supported a single connection on TLS. This is not flexible enough for complex testing scenarios which might require multiple simultaneous connections on different protocols.
Allow multiple simultaneous connections and add plain sockets as a protocol option. Rename the functions used for server scripting to hrnServerScript*() to make it clear they are related. Improve error messages when less input is received by the server than expected.
Also, do a bit of cleanup and add more comments.
Following up on 111d33c, implement the new interfaces for socket client/session. Now HTTP objects can be used over TLS or plain sockets.
This required adding ioSessionFd() and ioSessionRole() to provide the functionality of sckSessionFd() and sckSessionType(). sckClientHost() and sckClientPort don't make sense in a generic interface so they were replaced with ioSessionName().
Rather than calling storageS3New() directly, create the storage by loading a configuration and calling repoStorageGet(). This is a better end-to-end test and cuts down on a lot of redundant tests.
Add tests that include security tokens in error messages to ensure they are redacted.
Move sckSessionReadyRead()/Write() into the IoRead/IoWrite interfaces. This is a more logical place for them and the alternative would be to add them to the IoSession interface, which does not seem like a good idea.
This is mostly a refactor, but a big change is the select() logic in fdRead.c has been replaced by ioReadReady(). This was duplicated code that was being used by our protocol but not TLS. Since we have not had any problems with requiring poll() in the field this seems like a good time to remove our dependence on select().
Also, IoFdWrite now requires a timeout so update where required, mostly in the tests.
These interfaces allow the HttpClient and HttpSession objects to work with protocols other than TLS, .e.g. plain sockets. This is necessary to allow standard HTTP -- right now only HTTPS is allowed, i.e. HTTP over TLS.
For now only TlsClient and TlsSession have been converted to the new interfaces. SocketClient and SocketSession will also need to be converted but first sckSessionReadyRead() and sckSessionReadyWrite() need to be moved into the IoRead and IoWrite interfaces, since they are not a good fit for IoSession.
Pretty much everywhere handle is used what is really meant is file descriptor (fd). This terminology got migrated over from Perl and is just not quite correct, or at least not as correct as fd.
There were also plenty of places fd was used so now all uses are consistent.
The Perl code was not updated but might be in a future commit.
This does not appear to have been used in quite some time and the tests are equally useless because they don't prove the correct port was passed to httpClientNew().
Before 9f2d647 TLS errors included additional details in at least some cases. After 9f2d647 a connection to an HTTP server threw `TLS error [1]` instead of `unable to negotiate TLS connection: [336031996] unknown protocol`.
Bring back the detailed messages to make debugging TLS errors easier. Since the error routine is now generic the `unable to negotiate TLS connection context` is not available so the error looks like `TLS error [1:336031996] unknown protocol`.
This loop was using a lot of memory without freeing it at intervals.
Rewrite to use char arrays when possible to reduce memory that needs to be allocated and freed.
Zigzag encoding places the sign bit in the least significant bit so that -1 is encoded as 1, 1 as 2, etc. This moves as many bits as possible into the low order bits which is good for other types of encoding, e.g. base-128.
See https://en.wikipedia.org/wiki/Variable-length_quantity#Zigzag_encoding.
It seems like overkill to encode this when other enums (e.g. StorageInfoLevel) are passed as integers.
Instead note that StorageType values should not be changed and remove the special encoding.
The fix for = characters in info files (039d314) added JSON validation but discarded the resulting Variant which means the JSON is being parsed twice. This nearly doubles the time to load a manifest since a lot of complex JSON is involved.
Time to load a million file manifest:
Before 039d314: 7.8s
039d314: 15.5s
This patch: 7.5s
To fix this regression return the Variant in the callback so the caller does not have to parse it again. The new code appears slightly more efficient overall, probably because there are fewer operations against Strings.
We use the Z suffix in many functions to indicate that we are expecting a zero-terminated string so make this function conform to the pattern.
As a bonus the new name is a bit shorter, which is a good quality in a commonly-used function.
The manifest uses the = character as the key/value separator so = characters in the key cause parsing errors and lead to an error or segfault.
Since the value must be valid JSON we can keep checking the value on the right side of the = and stop building the key when the value is valid. It's a bit hackish but it does seem to do the job without breaking the manifest format.
Unsurprisingly this makes parsing about 50% slower but it's still more than fast enough. Parsing 10 million key/values takes about 6.5s for the old code and 10s for the new code. Since the value is used as JSON downstream we can reclaim most of this time by just passing the JSON value rather than making the callback reparse it. We'll save that for another commit, though.
Since the command has completed it is counterproductive to throw an error but still warn to indicate that something unusual happened.
Also fix the related issue that the local processes were not being shut down when they completed, which meant that they might timeout before being closed when pgbackrest terminated.
Use a test storage driver to allow manifestNewBuild() to be run against a test cluster at any scale without having to write files to disk.
Simplify the test by using the output of manifestNewBuild() to feed manifestSave() and manifestNewLoad().
Also add manifest size to the output.
Calculates the memory used by the context and all child contexts.
This is primarily useful for debugging but it is not conditional on DEBUG because it is useful for profile/performance tests.
A number of tests used invalid JSON values where an error was expected or the value would be ignored.
Update these tests to use valid JSON values so all values in the file can be validated even if they are not used.
Something like 3="string" would return an Int64 variant and ignore the invalid portion after the integer. Other JSON interface functions have this check but it was forgotten here.
There are no current issues because of this but we want to be able to validate arbitrary JSON strings and this function was not working correctly for that usage.
This function is not used in the core code so remove it and update the test where it was used.
There may eventually be a need for a strLstNewP() function but it doesn't seem worth the code churn until there is an actual requirement.
The old constructor was left around to reduce code churn during the migration but it just makes the code harder to read and search.
Remove the old constructor and rename all remaining instances to lstNewP(), which by default has the same semantics.
Testing against static checksums is valuable but it can be become burdensome when supporting multiple architectures.
Reduce the number of tests we are doing against static checksums when the architecture can cause the checksum to vary.
Bug Fixes:
* Fix restore --force acting like --force --delta. This caused restore to replace files based on timestamp and size rather than overwriting, which meant some files that should have been updated were left unchanged. Normal restore and restore --delta were not affected by this issue. (Reviewed by Cynthia Shang.)
Features:
* Azure support for repository storage. (Reviewed by Cynthia Shang, Don Seiler.)
* Add expire-auto option. This allows automatic expiration after a successful backup to be disabled. (Contributed by Stefan Fercot. Reviewed by Cynthia Shang, David Steele.)
Improvements:
* Asynchronous S3 multipart upload. (Reviewed by Stephen Frost.)
* Automatic retry for backup, restore, archive-get, and archive-push. (Reviewed by Cynthia Shang.)
* Disable query parallelism in PostgreSQL sessions used for backup control. (Reviewed by Stefan Fercot.)
* PostgreSQL 13 beta2 support. Changes to the control/catalog/WAL versions in subsequent betas may break compatibility but pgBackRest will be updated with each release to keep pace.
* Improve handling of invalid HTTP response status. (Reviewed by Cynthia Shang.)
* Improve error when pg1-path option missing for archive-get command. (Reviewed by Cynthia Shang.)
* Add hint when checksum delta is enabled after a timeline switch. (Reviewed by Matt Bunter, Cynthia Shang.)
* Use PostgreSQL instead of postmaster where appropriate. (Reviewed by Cynthia Shang.)
Documentation Bug Fixes:
* Fix incorrect example for repo-retention-full-type option. (Reported by Höseyin Sönmez.)
* Remove internal commands from HTML and man command references. (Reported by Cynthia Shang.)
Documentation Improvements:
* Update PostgreSQL versions used to build user guides. Also add version ranges to indicate that a user guide is accurate for a range of PostgreSQL versions even if it was built for a specific version. (Reviewed by Stephen Frost.)
* Update FAQ for expiring a specific backup set. (Contributed by Cynthia Shang. Reviewed by David Steele.)
* Update FAQ to clarify default PITR behavior. (Contributed by Cynthia Shang. Reviewed by David Steele.)
Remove all check and stanza-* tests except for the ones that are intended to succeed. The successful tests show that the queries run with expected results against each version of PG which should also validate queries for the failure tests in the unit tests.
Also remove the tests for --no-online backups since they don't require a database and are well tested in the unit tests.
The prior code was only able to use the main passphrase automatically and expected sub passphrases to be specified for each operation. This was fine for testing but hardly sufficient for a user-facing feature.
Update the code to determine which passphrase to use for any file in the repository and error when an invalid file or location is selected.
The repo-get command is still internal for now, but with this improvement it should be ready to be made public.
There are a few non version specific tests that need to be run in integration because we can't get coverage in the unit tests.
To save some time we'll only run those tests against the same version we use for expect testing.
If a local command, e.g. backupFile(), fails it will stop the entire process. Instead, retry local commands to deal with transient errors.
Remove special logic in the S3 storage driver to retry RequestTimeTooSkewed errors since this is now handled by the general retry mechanism in the places where it is most likely to happen, i.e. file read/write. Also, this error should have been entirely eliminated by the asynchronous TLS implementation.
The Azure storage driver exposes secrets in the query when using SAS authorization. These secrets can show up during logging or when an error occurs.
Allow redaction of queries to prevent secrets from being exposed in logs and errors.
A shared access signature (SAS) provides granular, delegated access to resources in a storage account. This is often preferable to using a shared key which provides more access and is a greater security risk if compromised.
Rework size limits so that this->size is always the current size no matter how much is allocated.
Most importantly, this removes the conditional in bufSize(), which makes it a better candidate for inlining.
This caused restore to replace files based on timestamp and size rather than overwriting, which meant some files that should have been updated were left unchanged. Normal restore and restore --delta were not affected by this issue.
httpUriDecode() reverses the encoding in httpUriEncode().
httpQueryNewStr() creates a new HttpQuery by parsing a query string.
httpQueryMerge() merges the contents of one query into another query.
Azure and Azure-compatible object stores can now be used for repository storage.
Currently only shared key authentication is supported but SAS will be added soon.
There don't appear to be any behavioral changes since PostgreSQL 12 and all the tests pass.
Changes to the control/catalog/WAL versions in subsequent betas may break compatibility but pgBackRest will be updated with each release to keep pace.
There is no need to have parallelism enabled in a backup control session. In particular, 9.6 marks pg_stop_backup() as parallel-safe but an error will be thrown if pg_stop_backup() is run in a worker.
When uploading large files the upload is split into multiple parts which are assembled at the end to create the final file. Previously we waited until each part was acknowledged before starting on the processing (i.e. compression, etc.) of the next part.
Now, the request for each part is sent while processing continues and the response is read just before sending the request for the next part. This asynchronous method allows us to continue processing while the S3 server formulates a response.
Testing from outside AWS in a high-bandwidth, low-latency environment showed a 35% improvement in the upload time of 1GB files. The time spent waiting for multipart notifications was reduced by ~300% (this measurement included the final part which is not uploaded asynchronously).
There are still some possible improvements: 1) the creation of the multipart id could be made asynchronous when it looks like the upload will need to be multipart (this may incur cost if the upload turns out not to be multipart). 2) allow more than one async request (this will use more memory).
A fair amount of refactoring was required to make the HTTP responses asynchronous. This may seem like overkill but having well-defined request, response, and session objects will also be advantageous for the upcoming HTTP server functionality.
Another advantage is that the lifecycle of an HttpSession is better defined. We only want to reuse sessions that complete the request/response cycle successfully, otherwise we consider the session to be in a bad state and would prefer to start clean with a new one. Previously, this required complex notifications to mark a session as "successfully done". Now, ownership of the session is passed to the request and then the response and only returned to the client after a successful response. If an error occurs anywhere along the way the session will be automatically closed by the object destructor when the request/response object is freed (depending on which one currently owns the session).
strCat() did not follow our convention of appending Z to functions that accept zero-terminated strings rather than String objects.
Add strCatZ() to accept zero-terminated strings and update strCat() to accept String objects.
Use LF_STR where appropriate but don't use other String constants because they do not improve readability.
Test matrices were previously simplified for the mock/* tests (e.g. d4410611, d489eb87) but not for real/all since the rules for which tests would run with which options was extremely complex. This only got more complex when new compression formats were added.
Because the loop-generated matrix was so large, mosts tests were skipped for most option combinations following arcane logic which was nearly impossible to decipher even when reading the code, and completely impossible from the test.pl interface. As a consequence, important tests got excluded. For example, backup from standby was excluded for most versions of PostgreSQL because it was only run once per distro, against the latest version to be included in that distro.
Simplify the tests by having a single run per PostgreSQL version and vary test parameters according to the capabilities of each version and the underlying distro. So, ZST testing is based on whether the distro supports ZST. Every test is run for each set of parameters based on the capabilities of the PostgreSQL version, e.g. backup from standby is not attempted on versions that don't support it.
Note that since more tests are running the overall time to run the mock/all tests has increased by about 20-25%. Some time may be saved my removing tests that are adequately covered by unit tests but that should the subject of another commit. Another option would be to limit some non version-specific tests to a single, well defined version of PostgreSQL, .e.g the version that is run by expect tests, currently 9.6.
The motivation for this refactor is that new storage drivers are coming and the loop-generated test matrix simply was not up to the task of adding them.
The following is an example of the new test log (note longer runtime of each test):
module=real, test=all, run=1, pg-version=10 (106.91s)
module=real, test=all, run=1, pg-version=9.5 (151.09s)
module=real, test=all, run=1, pg-version=9.2 (123.11s)
module=real, test=all, run=1, pg-version=9.1 (129s)
vs. the old test log (sub-second tests were skipped entirely):
module=real, test=all, run=2, pg-version=10 (0.31s)
module=real, test=all, run=3, pg-version=10 (0.26s)
module=real, test=all, run=4, pg-version=10 (60.39s)
module=real, test=all, run=1, pg-version=10 (69.12s)
module=real, test=all, run=6, pg-version=10 (34s)
module=real, test=all, run=5, pg-version=10 (42.75s)
module=real, test=all, run=2, pg-version=9.5 (0.21s)
module=real, test=all, run=3, pg-version=9.5 (0.21s)
module=real, test=all, run=4, pg-version=9.5 (0.21s)
module=real, test=all, run=5, pg-version=9.5 (0.26s)
module=real, test=all, run=6, pg-version=9.5 (0.21s)
module=real, test=all, run=1, pg-version=9.2 (72.78s)
module=real, test=all, run=2, pg-version=9.2 (0.26s)
module=real, test=all, run=3, pg-version=9.2 (0.31s)
module=real, test=all, run=4, pg-version=9.2 (0.21s)
module=real, test=all, run=5, pg-version=9.2 (0.21s)
module=real, test=all, run=6, pg-version=9.2 (0.21s)
module=real, test=all, run=1, pg-version=9.5 (88.41s)
module=real, test=all, run=2, pg-version=9.1 (0.21s)
module=real, test=all, run=3, pg-version=9.1 (0.26s)
module=real, test=all, run=4, pg-version=9.1 (0.21s)
module=real, test=all, run=5, pg-version=9.1 (0.31s)
module=real, test=all, run=6, pg-version=9.1 (0.26s)
module=real, test=all, run=1, pg-version=9.1 (72.4s)
The unsupported version error is showing up on older versions of PostgreSQL (e.g. 9.1, 9.2) on RHEL6 when setting up a standby with streaming replication. The error occurs when a client does not properly send a version number and it's not clear why it is happening here, but it does not appear to have anything to do with pgBackRest and only affects RHEL6, i.e. 9.1 and 9.2 do not show this error on other distros.
For now ignore the error since RHEL6 is nearly EOL.
HTTP is an acronym so it should be capitalized. Coding conventions dictate otherwise for function and type names but that should not have been propagated to comments and messages.
strPtr() is called more than any other function and during profiling (with or without optimization) it can end up using a disproportionate amount of the total runtime. Even though it is fast, the profiler has a minimum resolution for each function call so strPtr() will often end up towards the top of the list even though the real runtime is quite small.
Instead, inline strPtr() and indicate to gcc that it should be inlined even for non-optimized builds, since that's how profiles are usually generated.
To make strPtr() smaller require "this" to be non-NULL and add another function, strPtrNull(), to deal with the few cases where we need NULL handling.
As a bonus this makes the executable about 1% smaller even when compared to a prior optimized build which would inline some percentage of strPtr() calls.
There is no sense in generating detailed coverage reports in CI environments where they will never be seen. It takes time and format differences in some older versions can cause problems in the report generation code.
Note that missing coverage will still be reported on stdout and the test will fail.
These were missed in d41eea68 when the functionality of TEST_RESULT_STR() was changed. Using TEST_RESULT_STR() instead of TEST_RESULT_PTR() is more type-safe and clearer.
Add a comment to make it clear that TEST_RESULT_PTR() should be used only when a better alternative is not available.
This aligns better with general PostgreSQL usage and our own documentation (updated in 4bcef702).
Usage in the backup.manifest tests has not been updated since it might break the file format.
Expressions only worked at the first level of recursion because the expression was also being applied to paths so the path had to match the filter in order to recurse.
This is not considered a bug since it does not affect any existing code paths, but it is required for the general-purpose repo-ls command.
A truncated HTTP response status could lead to an an unfriendly error message, which would be retried, but could be confusing if the error was persistent and required debugging.
Improve the error handling overall to catch more error cases explicitly and respond better to edge cases.
Also update the terminology in comments to align with the RFC. Variable and function names were not changed because a refactor is intended for HTTP response and it doesn't seem worth the additional code churn.
The prior harness required a separate function to contain the server behavior but this made keeping the client/server code in sync very difficult and in general meant test writing took longer.
Now, commands to define server behavior are inline with the client code, which should greatly simplify test writing.
Bug Fixes:
* Fix issue checking if file links are contained in path links. (Reviewed by Cynthia Shang. Reported by Christophe Cavallié.)
* Allow pg-path1 to be optional for synchronous archive-push. (Reviewed by Cynthia Shang. Reported by Jerome Peng.)
* The expire command now checks if a stop file is present. (Fixed by Cynthia Shang. Reviewed by David Steele.)
* Handle missing reason phrase in HTTP response. (Reviewed by Cynthia Shang. Reported by Tenuun.)
* Increase buffer size for lz4 compression flush. (Reviewed by Cynthia Shang. Reported by Eric Radman.)
* Ignore pg-host* and repo-host* options for the remote command. (Reviewed by Cynthia Shang. Reported by Pavel Suderevsky.)
* Fix possibly missing pg1-* options for the remote command. (Reviewed by Cynthia Shang. Reported by Andrew L'Ecuyer.)
Features:
* Time-based retention for full backups. The --repo-retention-full-type option allows retention of full backups based on a time period, specified in days. (Contributed by Cynthia Shang, Pierre Ducroquet. Reviewed by David Steele.)
* Ad hoc backup expiration. Allow the user to remove a specified backup regardless of retention settings. (Contributed by Cynthia Shang. Reviewed by David Steele.)
* Zstandard compression support. Note that setting compress-type=zst will make new backups and archive incompatible (unrestorable) with prior versions of pgBackRest. (Reviewed by Cynthia Shang.)
* bzip2 compression support. Note that setting compress-type=bz2 will make new backups and archive incompatible (unrestorable) with prior versions of pgBackRest. (Contributed by Stephen Frost. Reviewed by David Steele, Cynthia Shang.)
* Add backup/expire running status to the info command. (Contributed by Stefan Fercot. Reviewed by David Steele.)
Improvements:
* Expire WAL archive only when repo-retention-archive threshold is met. WAL prior to the first full backup was previously expired after the first full backup. Now it is preserved according to retention settings. (Contributed by Cynthia Shang. Reviewed by David Steele.)
* Add local MD5 implementation so S3 works when FIPS is enabled. (Reviewed by Cynthia Shang, Stephen Frost. Suggested by Brian Almeida, John Kelley.)
* PostgreSQL 13 beta1 support. Changes to the control/catalog/WAL versions in subsequent betas may break compatibility but pgBackRest will be updated with each release to keep pace. (Reviewed by Cynthia Shang.)
* Reduce buffer-size default to 1MiB. (Reviewed by Stephen Frost.)
* Throw user-friendly error if expire is not run on repository host. (Contributed by Cynthia Shang. Reviewed by David Steele.)
The purpose of the remote command is to get access to local resources, so a remote should never start another remote. However, this could happen if there were host settings on the remote host, which ended badly with lock errors, loops, etc.
Add pg-local and repo-local options to indicate that the resource is local even if there are host settings.
Note that for the time being these options are internal and not intended for general usage. However, this is likely the direction needed to allow for more symmetric and manageable configurations.
Some pg1-* options are required by the remote so if they are not provided in the remote's configuration file then it may cause a configuration error, depending on the operation. This currently only applies to the pg1-path option.
This is still an issue for repo-* options but the same solution cannot be applied because some repo-* options are secure and cannot be passed on the command-line.
There don't appear to be any behavioral changes since PostgreSQL 12 and all the tests pass.
Changes to the control/catalog/WAL versions in subsequent betas may break compatibility but pgBackRest will be updated with each release to keep pace.
S3 requires the Content-MD5 header for many requests but MD5 is not available via OpenSSL when FIPS is enabled because it is considered to be insecure.
Even though our usage does not present any security risks a local M5 implementation is required to circumvent the over-broad FIPS restriction.
Vendorize the MD5 implementation found at https://openwall.info/wiki/people/solar/software/public-domain-source-code/md5 and add full coverage for the module in the common/crypto unit tests.
The prior default was determined by benchmarking the Perl code prior to the 1.0 release. In general buffer allocation was more expensive in Perl so large buffers gave the best performance. This was due to multiple buffer allocations for each filter in an IO operation.
The C code allocates fixed buffers for each IO operation so the cost for buffer allocation is lower than Perl. That being the case it made sense to benchmark the C code to determine the optimal buffer default.
The performance/storage tests were used to measure the performance of a variety of filters. 1GiB of data was processed by each filter 10 times and the results of the tests were averaged.
While most buffer sizes gave similar performance, 1MiB appeared to perform the best overall. Of course, different architectures are likely to yield different results but this seems like a sensible default. The buffer-size option may still need to be manually configured to give optimal results.
Raw test data for reference:
4MB buffer (prior default)
copy time 1807ms, avg time 180ms, avg throughput: 5942MB/s
md5 time 14200ms, avg time 1420ms, avg throughput: 756MB/s
sha1 time 11431ms, avg time 1143ms, avg throughput: 939MB/s
sha256 time 23463ms, avg time 2346ms, avg throughput: 457MB/s
gzip -6 time 381199ms, avg time 38119ms, avg throughput: 28MB/s
lz4 -1 time 15484ms, avg time 1548ms, avg throughput: 693MB/s
1MB buffer (new default)
copy time 1760ms, avg time 176ms, avg throughput: 6100MB/s
md5 time 13739ms, avg time 1373ms, avg throughput: 781MB/s
sha1 time 11025ms, avg time 1102ms, avg throughput: 973MB/s
sha256 time 22539ms, avg time 2253ms, avg throughput: 476MB/s
gzip -6 time 372995ms, avg time 37299ms, avg throughput: 28MB/s
lz4 -1 time 15118ms, avg time 1511ms, avg throughput: 710MB/s
512K buffer
copy time 1782ms, avg time 178ms, avg throughput: 6025MB/s
md5 time 13724ms, avg time 1372ms, avg throughput: 782MB/s
sha1 time 10959ms, avg time 1095ms, avg throughput: 979MB/s
sha256 time 22982ms, avg time 2298ms, avg throughput: 467MB/s
gzip -6 time 378120ms, avg time 37812ms, avg throughput: 28MB/s
lz4 -1 time 15484ms, avg time 1548ms, avg throughput: 693MB/s
256K buffer
copy time 1805ms, avg time 180ms, avg throughput: 5948MB/s
md5 time 13706ms, avg time 1370ms, avg throughput: 783MB/s
sha1 time 11074ms, avg time 1107ms, avg throughput: 969MB/s
sha256 time 22588ms, avg time 2258ms, avg throughput: 475MB/s
gzip -6 time 372645ms, avg time 37264ms, avg throughput: 28MB/s
lz4 -1 time 16346ms, avg time 1634ms, avg throughput: 656MB/s
Improve the accuracy of the calculations in several areas with better integer expressions.
Make the input buffer size configurable. Previously it was always 1mb, i.e. block size.
Use a macro for output results to reduce code duplication.
Reason phrases (e.g. OK) are optional in HTTP 1.1 but the space after the status code is not. When the reason phrase was missing the required space was trimmed along with the trailing CR leading to a format error.
Rework the logic to preserve the space and allow empty reason phrases.
Found while testing against the Backblaze S3-compatible API.
Vendorized code is copied from another project when a library is not available and a git subproject won't work. Currently all the vendorized code is copied from PostgreSQL but it makes sense to have a more general mechanism for indicating vendorized code.
The .vendor extension will be used to denote vendorized code in the same way that .auto is used to denote auto-generated code.
Rather than bS3 use strStorage which can indicate more than two storage types.
For the moment there are still only two storage types but this change is required before more can be added.
cdebfb09 added relative times to backup.into but a subtle issue was introduced that would cause the tests to fail if the time acquired by cmdExpire() was exactly the same as timeNow used to format backup.info. cmdExpire() was working correctly given the inputs, but the tests did not run predictably.
This was found while running the tests with --no-valgrind --no-coverage which allows them to run a lot faster, thus exposing the timing issue.
These tests required sudo to achieve complete coverage.
Add a new coverage exception, vm_covered, that applies to code that can only be covered in a container. When the test is run outside of a container code sections that require a container will be excluded with TEST_CONTAINER_REQUIRED and the coverage exception will be added to prevent a coverage error.
This does require marking up the core code with vm_covered, which in some modules (e.g. common/io/tls/client) can be extensive. It's possible that some of these tests can be rewritten to be less dependent on sudo but no attempt was made to do that here.
Only allow coverage summaries in a vm since coverage summaries outside a vm will not be complete, which was true even before this commit.
Update error types throw by bzip2 to be more consistent with gzip.
Update the bzip2 and gzip error default to be AssertError as that's the more common case in both, and add a 'break;' to the default clause -- we don't intend to be just falling through those case statements, even if the default is the last, we should be explicit about that.
Clean up some tabs that snuck in, rename a variable to be more clear, and add some comments.
The --repo-retention-full-type option allows retention of full backups based on a time period, specified in days.
The new option will default to 'count' and therefore will not affect current installations. Setting repo-retention-full-type to 'time' will allow the user to use a time period, in days, to indicate full backup retention. Using this method, a full backup can be expired only if the time the backup completed is older than the number of days set with repo-retention-full (calculated from the moment the 'expire' command is run) and at least one full backup meets the retention period. If archive retention has not been configured, then the default settings will expire archives that are prior to the oldest retained full backup. For example, if there are three full backups ending in times that are 25 days old (F1), 20 days old (F2) and 10 days old (F3), then if the full retention period is 15 days, then only F1 will be expired; F2 will be retained because F1 is not at least 15 days old.
Newer versions of sudo output this message to stderr when run in a container:
sudo: setrlimit(RLIMIT_CORE): Operation not permitted
See https://github.com/sudo-project/sudo/issues/42 for details.
A simple workaround is to prevent sudo from disabling core dumps. This seems safe enough because if sudo is segfaulting then core files are the least of our worries.
There are a number of Valgrind errors on Ubuntu 12.04 which do not happen on newer distro versions. However, suppressions for these errors have masked legitimate issues in subsequent code.
Instead, make suppressions VM specific so errors in other VMs are not masked.
Resolving localhost can vary based on the local network configuration so it is safer to just use a static IP.
This was found while testing on Travis-CI arm64.
bzip2 is a widely available, high-quality data compressor. It typically compresses files to within 10% to 15% of the best available techniques (the PPM family of statistical compressors), while being around twice as fast at compression and six times faster at decompression.
bzip2 is currently available on all supported platforms.
This was an oversight in 438b957f which added multiple compression type support. The booleans were interpreted as none and gz which works fine for the CompressType enum until the position of gz or none changes.
Zstandard is a fast lossless compression algorithm targeting real-time compression scenarios at zlib-level and better compression ratios. It's backed by a very fast entropy stage, provided by Huff0 and FSE library.
Zstandard version >= 1.0 is required, which is generally only available on newer distributions.
If the WAL path is absolute then pg1-path should be optional but in fact it was required to load pg_control.
Skip the pg_control check when pg1-path is not specified. The check against the stanza version/system-id remains to protect the repo from corruption.
Perhaps this was intended to verify the WAL size but was never implemented.
Verifying the WAL size is probably a good idea so this member may be added back if the feature is implemented.
An upcoming feature requires new parameters for storagePosixNew() and this causes a lot of churn because almost every test creates a Posix storage object. Some refactoring in the tests might reduce this duplication but storagePosixNew() is collecting a lot of parameters so converting to storagePosixNewP() makes sense in any case.
There are relatively few call sites in the core code but they still benefit from better readability after this change.
There is no conflict if the path containing a file link is a parent path of a path link. The Perl code apparently had this right but the migration to C missed it.
Exclude this case when checking for link conflicts.
There have been a number of segfaults reported because a string option expected to be non-null was actually null. This is generally due to options that are expected to be set but are in fact optional.
Protect against this by creating cfgOptionStrNull() to get options that can be null, while changing cfgOptionStr() to always expect non-null. There are relatively few places where nulls are expected.
There is definitely a chance for breakage here as null options might currently be working in the field but will be caught by this new check. Hopefully introducing the check early in the release cycle will allow us to catch any issues.
It makes sense to do this check right after the first compression so any issues are caught early.
Also, none of the current compression formats omit decompressCmd so make the test mandatory.
Previously when retention-archive was set (either by the user or by default), archives prior to the archive-start of the oldest remaining full backup (after backup expiration occurred) would be expired even though the retention-archive threshold had not been met. For example, if there were 1 full backup remaining after backup expiration and the retention-archive was set to 2 and retention-archive-type=full, then archives prior to the archive-start of the remaining full backup would still be removed even though retention-archive required 2 full backups remaining before archives should be expired.
The thought was to keep the archive directory clean and since the full backup did not require prior archives, it was safe to delete them. However, this has caused problems for some users in the past (because they needed the WAL for other purposes) and with the new adhoc and time-based retention features, it was decided that the archives should remain until the threshold was met. The archives will eventually be removed and if having them causes space issues, the expire command and the retention-archive can always be run and adjusted.
Coverity was concerned that regExpError() might return and lead to an invalid reference of "this". This was unlikely since the function should never return but Coverity didn't know that. Also, a difference in error-handling logic at the two sites could cause the issue Coverity reported if they were to get out of sync.
Fix by refactoring out the core error function so that it is clear it will never return.
If an option may not be valid for a command it should be checked with cfgOptionValid() or cfgOptionTest().
It appears this rule is followed pretty strictly since the only changes required were in unit tests.
The specified backup set (i.e. the backup label provided and all of its dependent backups, if any) will be expired regardless of backup retention rules except that at least one full backup must remain in the repository.
Each option type enforced its own constraints but there was a lot of duplication. Centralize the enforcement to remove the duplication.
Also convert the option type assert to a production error. This is unlikely to happen in production but the test is quite cheap so it can't hurt.
Finally, add a NULL check. Most option types can never be NULL.
This is implemented by checking for a backup lock on the host where info is running so there are a few limitations:
* It is not currently possible to know which command is running: backup, expire, or stanza-*. The stanza commands are very unlikely to be running so it's pretty safe to guess backup/expire. Command information may be added to the lock file to improve the accuracy of the reported command.
* If the info command is run on a host that is not participating in the backup, e.g. a standby, then there will be no backup lock. This seems like a minor limitation since running info on the repo or primary host is preferred.
Make the restore clean process look more like manifest build, i.e. do cleanup of each target root directory outside the main cleanup callback. This means some code duplication but removes the logic handling "dot" paths.
Add tests for both restore and backup (which already worked but was not tested).
Bug Fixes:
* Remove empty subexpression from manifest regular expression. MacOS was not happy about this though other platforms seemed to work fine. (Fixed by David Raftis.)
Improvements:
* Non-blocking TLS implementation. (Reviewed by Slava Moudry, Cynthia Shang, Stephen Frost.)
* Only limit backup copy size for WAL-logged files. The prior behavior could possibly lead to postgresql.conf or postgresql.auto.conf being truncated in the backup. (Reviewed by Cynthia Shang.)
* TCP keep-alive options are configurable. (Suggested by Marc Cousin.)
* Add io-timeout option.
Timeout used for connections and read/write operations.
Note that the entire read/write operation does not need to complete within this timeout but some progress must be made, even if it is only a single byte.
The prior blocking implementation seemed to be prone to locking up on some (especially recent) kernel versions. Since we were unable to reproduce the issue in a development environment we can only speculate as to the cause, but there is a good chance that blocking sockets were the issue or contributed to the issue.
So move to a non-blocking implementation to hopefully clear up these issues. Testing in production environments that were prone to locking shows that the approach is promising and at the very least not a regression.
The main differences from the blocking version are the non-blocking connect() implementation and handling of WANT_READ/WANT_WRITE retries for all SSL*() functions.
Timeouts in the tests needed to be increased because socket connect() and TLS SSL_connect() were not included in the timeout before. The tests don't run any slower, though. In fact, all platforms but Ubuntu 12.04 worked fine with the shorter timeouts.
select() is a bit old-fashioned and cumbersome to use. Since the select() code needed to be modified to handle write ready this seems like a good time to upgrade to poll().
poll() has been around for a long time so there doesn't seem to be any need to provide a fallback to select().
Also change the error on timeout from FileReadError to ProtocolError. This works better for read vs. write and failure to poll() is indicative of a protocol error or unexpected EOF.
The prior behavior introduced in dcddf3a5 could possibly lead to postgresql.conf or postgresql.auto.conf being truncated in the backup since they are copied via tmp files and could change size during the backup.
In general it seems safer to limit this feature to WAL-logged files which will be reconstructed during recovery.
Building on these platforms gives us better coverage for our build code. Cirrus CI was chosen because it is the only service that supports FreeBSD (that we could find).
The FreedBSD configuration for Vagrant is currently just enough to perform a build.
The MacOS configuration is not actually for Vagrant (yet) but does show the steps needed to setup the build environment on MacOS.
This allows us to add new configurations mostly without changing the behavior of vagrant from the command line, i.e. vagrant up and vagrant ssh will continue to bring up the default configuration.
However, vagrant destroy -f will remove all configurations. That's really only a change in behavior if more than one configuration is running, which is not currently possible.
A session looks much the same whether it is initiated from the client or the server, so use the session objects to implement the TLS, HTTP, and S3 test servers.
For TLS, at least, there are some differences between client and server sessions so add a client/server type to SocketSession to determine how the session was initiated.
Aside from reducing code duplication, the main advantage is that the test server will now timeout rather than hanging indefinitely when less input that expected is received.
Previously an error was only thrown when errno was set but in practice this is usually not the case. This may have something to do with getting errno late but attempts to get it earlier have not been successful. It appears that errno usually gets cleared and spot research seems to indicate that other users have similar issues.
An error at this point indicates unexpected EOF so it seems better to just throw an error all the time and be consistent.
To test this properly our test server needs to call SSL_shutdown() except when the client expects this error.
This abstraction allows the session code to be shared between the TLS client and (upcoming) server code.
Session management is no longer implemented in TlsClient so the HttpClient was updated to free and create sessions as needed. No test changes were required for HttpClient so the functionality should be unchanged.
Mechanical changes to the TLS tests were required to use TlsSession where appropriate rather than TlsClient. There should be no change in functionality other than how sessions are managed, i.e. using tlsClientOpen()/tlsSessionFree() rather than just tlsClientOpen().
The errorInternalThrowSys*() functions were marked as returning during coverage testing even when they had no possibility to return, i.e. the error parameter was set to constant true. This meant the compiler would treat the functions as returning even when they would not.
Instead create completely separate functions for coverage to use for THROW_ON_SYS_ERROR*() that can return and leave the regular functions marked __noreturn__.
This abstraction allows the session code to be shared between the socket client and (upcoming) server code. There should no difference in how the code works -- only the organization has changed. Note that no changes to the tests were required.
This same abstraction will be required for TlsClient but that will be done in a separate commit because it requires test changes.
These forks were done in a custom way (not sure why) and lack the capability of the standard macros for the parent to wait for child exit.
This mean that the server would continue to run after the tests were complete and that multiple servers could run at once. This caused subtle timing and connection issues that required larger timeouts to resolve.
Don't change the timeouts here since they need to be adjusted in future commits anyway.
It is pretty much impossible for a static IP to not resolve to an address but in theory the error could catch other conditions so it seems best to keep it.
When the Vagrant file was updated to use pgbackrest/ vs /backrest/ as the location for executing tests and building the documentation, parts of the contributing.xml (and hence the CONTRIBUTING.md) were not updated since some parts of the document are not actually executed when the CONTRIBUTING.md is built from contributing.xml: those parts that are executed were updated but those parts that are not executed were not.
This commit fixes the contributing.xml issue but also removes test/README.md as its contents were out of date and redundant given that they are covered in CONTRIBUTING.md.
