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.

Sign up for GitHub

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

Add support for compressions and make LZ4 default for backup 2.0 #833

Open
wants to merge 5 commits into
base: 3.x
Choose a base branch
from
Open

Add support for compressions and make LZ4 default for backup 2.0 #833

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
a551e7e
initial change - TTL
arunagrawal-84 May 6, 2019
ff65515
Removing empty DSE extension module
arunagrawal-84 May 6, 2019
a1082b4
Change to instant for object lock
arunagrawal-84 Jun 7, 2019
404714c
add lz4 and no compression type
arunagrawal-84 Jul 30, 2019
645c058
Pass directives to backup file system.
arunagrawal-84 Sep 20, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 8 additions & 1 deletion priam/src/main/java/com/netflix/priam/aws/S3FileSystem.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -213,7 +213,14 @@ protected long uploadFileImpl(Path localPath, Path remotePath) throws BackupRest
}
byte[] chunk = byteArrayOutputStream.toByteArray();
long compressedFileSize = chunk.length;
rateLimiter.acquire(chunk.length);

/**
* Weird, right that we are checking for length which is positive. You can thanks
* this to sometimes C* creating files which are zero bytes, and giving that in
* snapshot for some unknown reason.
*/
if (chunk.length > 0) rateLimiter.acquire(chunk.length);
Copy link
Contributor

Choose a reason for hiding this comment

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

I assume we still want to proceed with the upload of this 0 byte file to keep our s3 snapshots consistent with C* snapshot view, and hence we do not skip the rest of the upload code path even if the file has 0 byte - is this correct?


ObjectMetadata objectMetadata = getObjectMetadata(localPath);
objectMetadata.setContentLength(chunk.length);
PutObjectRequest putObjectRequest =
Expand Down
5 changes: 4 additions & 1 deletion priam/src/main/java/com/netflix/priam/backup/AbstractFileSystem.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,7 +80,10 @@ public AbstractFileSystem(
// Add notifications.
this.addObserver(backupNotificationMgr);
this.objectCache =
CacheBuilder.newBuilder().maximumSize(configuration.getBackupQueueSize()).build();
CacheBuilder.newBuilder()
.maximumSize(configuration.getBackupQueueSize())
.expireAfterAccess(configuration.getBackupCacheTTLInDays(), TimeUnit.DAYS)
.build();
tasksQueued = new ConcurrentHashMap<>().newKeySet();
/*
Note: We are using different queue for upload and download as with Backup V2.0 we might download all the meta
Expand Down
11 changes: 11 additions & 0 deletions priam/src/main/java/com/netflix/priam/config/IConfiguration.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -826,6 +826,17 @@ default int getBackupQueueSize() {
return 100000;
}

/**
* TTL (in days) for the entries - filenames which will be cached internally when successfully
* uploaded to remote file system. This cache is used to avoid bombarding the remote file system
* to check if object exists.
*
* @return
*/
default int getBackupCacheTTLInDays() {
return 2;
}

/**
* Queue size to be used for file downloads. Note that once queue is full, we would wait for
* {@link #getDownloadTimeout()} to add any new item before declining the request and throwing
Expand Down