Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Make backups openable as read-only DBs #8142

Closed
wants to merge 5 commits into from

Conversation

pdillinger
Copy link
Contributor

@pdillinger pdillinger commented Apr 1, 2021

Summary: A current limitation of backups is that you don't know the
exact database state of when the backup was taken. With this new
feature, you can at least inspect the backup's DB state without
restoring it by opening it as a read-only DB.

Rather than add something like OpenAsReadOnlyDB to the BackupEngine API,
which would inhibit opening stackable DB implementations read-only
(if/when their APIs support it), we instead provide a DB name and Env
that can be used to open as a read-only DB.

Possible follow-up work:

  • Add a version of GetBackupInfo for a single backup.
  • Let CreateNewBackup return the BackupID of the newly-created backup.

Implementation details:

Refactored ChrootFileSystem to split off new base class RemapFileSystem,
which allows more general remapping of files. We use this base class to
implement BackupEngineImpl::RemapSharedFileSystem.

To minimize API impact, I decided to just add these fields name_for_open
and env_for_open to those set by GetBackupInfo when
include_file_details=true. Creating the RemapSharedFileSystem adds a bit
to the memory consumption, perhaps unnecessarily in some cases, but this
has been mitigated by (a) only initialize the RemapSharedFileSystem
lazily when GetBackupInfo with include_file_details=true is called, and
(b) using the existing shared_ptr<FileInfo> objects to hold most of the
mapping data.

To enhance API safety, RemapSharedFileSystem is wrapped by new
ReadOnlyFileSystem which rejects any attempts to write. This uncovered a
couple of places in which DB::OpenForReadOnly would write to the
filesystem, so I fixed these. Added a release note because this affects
logging.

Additional minor refactoring in backupable_db.cc to support the new
functionality.

Test Plan: new test (run with ASAN and UBSAN), added to stress test and
ran it for a while with amplified backup_one_in

Summary: A current limitation of backups is that you don't know the
exact database state of when the backup was taken. With this new
feature, you can at least inspect the backup's DB state without
restoring it by opening it as a read-only DB.

Rather than add something like OpenAsReadOnlyDB to the BackupEngine API,
which would inhibit opening stackable DB implementations read-only
(if/when their APIs support it), we instead provide a DB name and Env
that can be used to open as a read-only DB.

Possible follow-up work:

* Add a version of GetBackupInfo for a single backup.
* Let CreateNewBackup return the BackupID of the newly-created backup.

Implementation details:

Refactored ChrootFileSystem to split off new base class RemapFileSystem,
which allows more general remapping of files. We use this base class to
implement BackupEngineImpl::RemapSharedFileSystem.

To minimize API impact, I decided to just add these fields name_for_open
and env_for_open to those set by GetBackupInfo when
include_file_details=true. Creating the RemapSharedFileSystem adds a bit
to the memory consumption, perhaps unnecessarily in some cases, but this
has been mitigated by (a) only initialize the RemapSharedFileSystem
lazily when GetBackupInfo with include_file_details=true is called, and
(b) using the existing shared_ptr<FileInfo> objects to hold most of the
mapping data.

To enhance API safety, RemapSharedFileSystem is wrapped by new
ReadOnlyFileSystem which rejects any attempts to write. This uncovered a
couple of places in which DB::OpenForReadOnly would write to the
filesystem, so I fixed these. Added a release note because this affects
logging.

Additional minor refactoring in backupable_db.cc to support the new
functionality.

Test Plan: new test (run with ASAN and UBSAN), added to stress test and
ran it for a while with amplified backup_one_in
@facebook-github-bot
Copy link
Contributor

@pdillinger has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

Comment on lines +868 to +870
if (read_only) {
db_id_ = env_->GenerateUniqueId();
return Status::OK();
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This seems wrong to me. The documentation says/implies that if I open the same DB twice, the DB ID should be the same. In this case, opening the same DB read-only will return different IDs, which seems wrong.

Also, if db_id_ is not empty, the code still call SetIdentityFile, which will attempt to write a file to the Env...

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't know the history on why, but backups do not save the IDENTITY file. If you restore to a fresh dir, you get a new db ID, which is what this replicates.

env/fs_remap.h Outdated
IODebugContext* dbg) override;
};

// An abstract FileSystem wrapper that only allows read-only operation.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is not abstract is it? One could create an instance of it.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thoughtless copy-paste error

env/fs_remap.h Outdated
Comment on lines 133 to 136
class ReadOnlyFileSystem : public FileSystemWrapper {
static inline IOStatus FailReadOnly() {
return IOStatus::IOError("Attempted write to ReadOnlyFileSystem");
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Since this is not really part of the fs_remap, can we move it to its own header file?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Agreed

env/fs_remap.h Outdated
// guarantees.
class ReadOnlyFileSystem : public FileSystemWrapper {
static inline IOStatus FailReadOnly() {
return IOStatus::IOError("Attempted write to ReadOnlyFileSystem");
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why an IOError and not NotSupported? Seems like an IOError could lead to situations where the operation might be retried when it will always fail.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I attempted to replicate the same error you would get by trying to write to a read-only filesystem. I believe this construction sets retryable to false.

Comment on lines +313 to +318
// This Env might or might not be shared with other backups. To work
// around DBOptions::env being a raw pointer, this is a shared_ptr so
// that keeping either this BackupInfo, the BackupEngine, or a copy of
// this shared_ptr alive is sufficient to keep the Env alive for use by
// a read-only DB.
std::shared_ptr<Env> env_for_open;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This seems very dangerous to me. If env_for_open is set to Env::Default (for example), then won't the default env be deleted once this is out of scope?

Cn this be a FileSystem instead of an Env? Are there operations in the Env that you require that are not in the FS?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

env_for_open is a field, so for an API user to reset it to some raw ptr they would have to use .reset() on a raw ptr they don't control the lifetime of. What's the new hazard here? Also this is an "output" struct for our API. Why would an API user be writing to these fields at all?

Cn this be a FileSystem instead of an Env?

Then the caller has unnecessary complexity of creating their own temporary Env. What's the benefit?

const std::string kMetaDirSlash = kMetaDirName + "/";
const std::string kSharedDirSlash = kSharedDirName + "/";
const std::string kSharedChecksumDirSlash = kSharedChecksumDirName + "/";

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can/Should these be static const?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

They're in an anonymous namespace. I didn't think it was important to put them in the class.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I was not suggesting in a class, just static to guarantee scoping and creation constraints.

Comment on lines +230 to +244
static inline std::string WithoutTrailingSlash(const std::string& path) {
if (path.empty() || path.back() != '/') {
return path;
} else {
return path.substr(path.size() - 1);
}
}

static inline std::string WithTrailingSlash(const std::string& path) {
if (path.empty() || path.back() != '/') {
return path + '/';
} else {
return path;
}
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Would NormalizePath meet the needs here?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It does not remove trailing /

Copy link
Contributor

@ajkr ajkr left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM! Glad validation can now be done natively without full restore. Seems good to push users through the DB open functions.

@@ -1404,7 +1405,8 @@ Status StressTest::TestBackupRestore(
}
}
}
if (s.ok()) {
if (s.ok() && !inplace_not_restore) {
// Purge early if restoring
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe comment that we do early purge in restore mode so we can cover reading from a restored+purged DB? It's probably obvious to a lot of people but I had to stop and think...

env/fs_remap.h Outdated
Comment on lines 133 to 136
class ReadOnlyFileSystem : public FileSystemWrapper {
static inline IOStatus FailReadOnly() {
return IOStatus::IOError("Attempted write to ReadOnlyFileSystem");
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Agreed

@facebook-github-bot
Copy link
Contributor

@pdillinger has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@pdillinger has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

@facebook-github-bot
Copy link
Contributor

@pdillinger merged this pull request in 879357f.

pdillinger added a commit to pdillinger/rocksdb that referenced this pull request Apr 7, 2021
Summary: Forgot to re-test crash test after adding read-only filesystem
enforcement to facebook#8142. The problem is ReadOnlyFileSystem would reject
CreateDirIfMissing whenever DBOptions::create_if_missing=true. The fix
that is better for users is to allow CreateDirIfMissing in
ReadOnlyFileSystem if the directory exists, so that they don't cause a
failure on using create_if_missing with opening backups as read-only
DBs.

Also fixed a couple of lints.

Test Plan: local blackbox_crash_test with amplified backup_one_in
facebook-github-bot pushed a commit that referenced this pull request Apr 7, 2021
Summary:
Forgot to re-test crash test after adding read-only filesystem
enforcement to #8142. The problem is ReadOnlyFileSystem would reject
CreateDirIfMissing whenever DBOptions::create_if_missing=true. The fix
that is better for users is to allow CreateDirIfMissing in
ReadOnlyFileSystem if the directory exists, so that they don't cause a
failure on using create_if_missing with opening backups as read-only
DBs. Added this option test to the unit test (in addition to being in the
crash test).

Also fixed a couple of lints.

And some better messaging from 'make format' so that when you run it
with uncommitted changes, it's clear that it's only checking the
uncommitted changes.

Pull Request resolved: #8161

Test Plan: local blackbox_crash_test with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27614409

Pulled By: pdillinger

fbshipit-source-id: 63ccb626c7e34c200d61c6bca2a8f60da9015179
levichen94 pushed a commit to bytedance/terarkdb that referenced this pull request Jun 18, 2021
Summary:
A current limitation of backups is that you don't know the
exact database state of when the backup was taken. With this new
feature, you can at least inspect the backup's DB state without
restoring it by opening it as a read-only DB.

Rather than add something like OpenAsReadOnlyDB to the BackupEngine API,
which would inhibit opening stackable DB implementations read-only
(if/when their APIs support it), we instead provide a DB name and Env
that can be used to open as a read-only DB.

Possible follow-up work:

* Add a version of GetBackupInfo for a single backup.
* Let CreateNewBackup return the BackupID of the newly-created backup.

Implementation details:

Refactored ChrootFileSystem to split off new base class RemapFileSystem,
which allows more general remapping of files. We use this base class to
implement BackupEngineImpl::RemapSharedFileSystem.

To minimize API impact, I decided to just add these fields `name_for_open`
and `env_for_open` to those set by GetBackupInfo when
include_file_details=true. Creating the RemapSharedFileSystem adds a bit
to the memory consumption, perhaps unnecessarily in some cases, but this
has been mitigated by (a) only initialize the RemapSharedFileSystem
lazily when GetBackupInfo with include_file_details=true is called, and
(b) using the existing `shared_ptr<FileInfo>` objects to hold most of the
mapping data.

To enhance API safety, RemapSharedFileSystem is wrapped by new
ReadOnlyFileSystem which rejects any attempts to write. This uncovered a
couple of places in which DB::OpenForReadOnly would write to the
filesystem, so I fixed these. Added a release note because this affects
logging.

Additional minor refactoring in backupable_db.cc to support the new
functionality.

Pull Request resolved: facebook/rocksdb#8142

Test Plan:
new test (run with ASAN and UBSAN), added to stress test and
ran it for a while with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27535408

Pulled By: pdillinger

fbshipit-source-id: 04666d310aa0261ef6b2385c43ca793ce1dfd148
Signed-off-by: Changlong Chen <levisonchen@live.cn>
levichen94 pushed a commit to bytedance/terarkdb that referenced this pull request Jun 18, 2021
Summary:
Forgot to re-test crash test after adding read-only filesystem
enforcement to facebook/rocksdb#8142. The problem is ReadOnlyFileSystem would reject
CreateDirIfMissing whenever DBOptions::create_if_missing=true. The fix
that is better for users is to allow CreateDirIfMissing in
ReadOnlyFileSystem if the directory exists, so that they don't cause a
failure on using create_if_missing with opening backups as read-only
DBs. Added this option test to the unit test (in addition to being in the
crash test).

Also fixed a couple of lints.

And some better messaging from 'make format' so that when you run it
with uncommitted changes, it's clear that it's only checking the
uncommitted changes.

Pull Request resolved: facebook/rocksdb#8161

Test Plan: local blackbox_crash_test with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27614409

Pulled By: pdillinger

fbshipit-source-id: 63ccb626c7e34c200d61c6bca2a8f60da9015179
Signed-off-by: Changlong Chen <levisonchen@live.cn>
mm304321141 pushed a commit to bytedance/terarkdb that referenced this pull request Jun 23, 2021
Summary:
A current limitation of backups is that you don't know the
exact database state of when the backup was taken. With this new
feature, you can at least inspect the backup's DB state without
restoring it by opening it as a read-only DB.

Rather than add something like OpenAsReadOnlyDB to the BackupEngine API,
which would inhibit opening stackable DB implementations read-only
(if/when their APIs support it), we instead provide a DB name and Env
that can be used to open as a read-only DB.

Possible follow-up work:

* Add a version of GetBackupInfo for a single backup.
* Let CreateNewBackup return the BackupID of the newly-created backup.

Implementation details:

Refactored ChrootFileSystem to split off new base class RemapFileSystem,
which allows more general remapping of files. We use this base class to
implement BackupEngineImpl::RemapSharedFileSystem.

To minimize API impact, I decided to just add these fields `name_for_open`
and `env_for_open` to those set by GetBackupInfo when
include_file_details=true. Creating the RemapSharedFileSystem adds a bit
to the memory consumption, perhaps unnecessarily in some cases, but this
has been mitigated by (a) only initialize the RemapSharedFileSystem
lazily when GetBackupInfo with include_file_details=true is called, and
(b) using the existing `shared_ptr<FileInfo>` objects to hold most of the
mapping data.

To enhance API safety, RemapSharedFileSystem is wrapped by new
ReadOnlyFileSystem which rejects any attempts to write. This uncovered a
couple of places in which DB::OpenForReadOnly would write to the
filesystem, so I fixed these. Added a release note because this affects
logging.

Additional minor refactoring in backupable_db.cc to support the new
functionality.

Pull Request resolved: facebook/rocksdb#8142

Test Plan:
new test (run with ASAN and UBSAN), added to stress test and
ran it for a while with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27535408

Pulled By: pdillinger

fbshipit-source-id: 04666d310aa0261ef6b2385c43ca793ce1dfd148
Signed-off-by: Changlong Chen <levisonchen@live.cn>
mm304321141 pushed a commit to bytedance/terarkdb that referenced this pull request Jun 23, 2021
Summary:
Forgot to re-test crash test after adding read-only filesystem
enforcement to facebook/rocksdb#8142. The problem is ReadOnlyFileSystem would reject
CreateDirIfMissing whenever DBOptions::create_if_missing=true. The fix
that is better for users is to allow CreateDirIfMissing in
ReadOnlyFileSystem if the directory exists, so that they don't cause a
failure on using create_if_missing with opening backups as read-only
DBs. Added this option test to the unit test (in addition to being in the
crash test).

Also fixed a couple of lints.

And some better messaging from 'make format' so that when you run it
with uncommitted changes, it's clear that it's only checking the
uncommitted changes.

Pull Request resolved: facebook/rocksdb#8161

Test Plan: local blackbox_crash_test with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27614409

Pulled By: pdillinger

fbshipit-source-id: 63ccb626c7e34c200d61c6bca2a8f60da9015179
Signed-off-by: Changlong Chen <levisonchen@live.cn>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

4 participants