inteligr8/alfresco-community-repo
1
0
Fork 0
mirror of https://github.com/Alfresco/alfresco-community-repo.git synced 2025-07-31 17:39:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

MNT-20770: Share non responsive during direct download from S3 if content store selector is also configured (#654)

Browse Source 
- add specific implementation for exists to bypass getting the complete reader and in some cases (e.g. aggregating a CachingContentStore) trigger a complete download of the file
This commit is contained in:
Cristian Turlica
2019-11-05 14:30:42 +02:00
committed by GitHub
parent 67e9e5fbb7
commit 933161a914
1 changed files with 225 additions and 154 deletions
Download Patch File Download Diff File Expand all files Collapse all files

379
src/main/java/org/alfresco/repo/content/replication/AggregatingContentStore.java
View File

@@ -23,8 +23,8 @@
* along with Alfresco. If not, see <http://www.gnu.org/licenses/>.
* #L%
*/
package org.alfresco.repo.content.replication;
package org.alfresco.repo.content.replication;
import java.util.List;
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReadWriteLock;
@@ -34,161 +34,232 @@ import org.alfresco.error.AlfrescoRuntimeException;
import org.alfresco.repo.content.AbstractContentStore;
import org.alfresco.repo.content.ContentContext;
import org.alfresco.repo.content.ContentStore;
import org.alfresco.repo.content.UnsupportedContentUrlException;
import org.alfresco.repo.content.caching.CachingContentStore;
import org.alfresco.service.cmr.repository.ContentIOException;
import org.alfresco.service.cmr.repository.ContentReader;
import org.alfresco.service.cmr.repository.ContentWriter;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
/**
* <h1><u>Aggregating Content Store</u></h1>
* <p>
* A content store implementation that aggregates a set of stores. Content is not
* persisted by this store, but rather it relies on any number of
* child {@link org.alfresco.repo.content.ContentStore stores} to provide access
* to content readers and writers.
* <p>
* The order in which the stores appear in the list of stores participating is
* important. The first store in the list is known as the <i>primary store</i>.
<p>
* Content is written to the primary store only. The other stores are
* only used to retrieve content and the primary store is not updated with
* the content.
*
* @author Derek Hulley
* @author Mark Rogers
* @see CachingContentStore
*/
public class AggregatingContentStore extends AbstractContentStore
{
private static Log logger = LogFactory.getLog(AggregatingContentStore.class);
private ContentStore primaryStore;
private List<ContentStore> secondaryStores;
private Lock readLock;
/**
* Default constructor
*/
public AggregatingContentStore()
{
ReadWriteLock storeLock = new ReentrantReadWriteLock();
readLock = storeLock.readLock();
}
/**
* Set the primary store that content will be replicated to or from
*
* @param primaryStore the primary content store
*/
public void setPrimaryStore(ContentStore primaryStore)
{
this.primaryStore = primaryStore;
}
/**
* Set the secondary stores that this component will replicate to or from
*
* @param secondaryStores a list of stores to replicate to or from
*/
public void setSecondaryStores(List<ContentStore> secondaryStores)
{
this.secondaryStores = secondaryStores;
}
/**
* @return Returns <tt>true</tt> if the primary store supports writing
*/
public boolean isWriteSupported()
{
return primaryStore.isWriteSupported();
}
/**
* @return Returns <tt>true</tt> if the primary store supports the URL
*/
@Override
public boolean isContentUrlSupported(String contentUrl)
{
return primaryStore.isContentUrlSupported(contentUrl);
}
/**
* @return Return the primary store root location
*/
@Override
public String getRootLocation()
{
return primaryStore.getRootLocation();
}
/**
* Forwards the call directly to the first store in the list of stores.
*/
public ContentReader getReader(String contentUrl) throws ContentIOException
{
if (primaryStore == null)
{
throw new AlfrescoRuntimeException("ReplicatingContentStore not initialised");
}
// get a read lock so that we are sure that no replication is underway
readLock.lock();
try
{
// get a reader from the primary store
ContentReader primaryReader = primaryStore.getReader(contentUrl);
// give it straight back if the content is there
if (primaryReader.exists())
{
return primaryReader;
}
// the content is not in the primary reader so we have to go looking for it
for (ContentStore store : secondaryStores)
{
ContentReader reader = store.getReader(contentUrl);
if (reader.exists())
{
// found the content in a secondary store
return reader;
}
}
return primaryReader;
}
finally
{
readLock.unlock();
}
}
public ContentWriter getWriter(ContentContext ctx)
{
// get the writer
ContentWriter writer = primaryStore.getWriter(ctx);
return writer;
}
/**
* Performs a delete on the local store and if outbound replication is on, propogates
* the delete to the other stores too.
*
* @return Returns the value returned by the delete on the primary store.
*/
public boolean delete(String contentUrl) throws ContentIOException
{
// delete on the primary store
boolean deleted = primaryStore.delete(contentUrl);
if (logger.isDebugEnabled())
{
logger.debug("Deleted content for URL: " + contentUrl);
}
return deleted;
}
}
/**
* <h1><u>Aggregating Content Store</u></h1>
* <p>
* A content store implementation that aggregates a set of stores. Content is not
* persisted by this store, but rather it relies on any number of
* child {@link org.alfresco.repo.content.ContentStore stores} to provide access
* to content readers and writers.
* <p>
* The order in which the stores appear in the list of stores participating is
* important. The first store in the list is known as the <i>primary store</i>.
<p>
* Content is written to the primary store only. The other stores are
* only used to retrieve content and the primary store is not updated with
* the content.
*
* @author Derek Hulley
* @author Mark Rogers
* @see CachingContentStore
*/
public class AggregatingContentStore extends AbstractContentStore
{
private static Log logger = LogFactory.getLog(AggregatingContentStore.class);
private ContentStore primaryStore;
private List<ContentStore> secondaryStores;
private Lock readLock;
/**
* Default constructor
*/
public AggregatingContentStore()
{
ReadWriteLock storeLock = new ReentrantReadWriteLock();
readLock = storeLock.readLock();
}
/**
* Set the primary store that content will be replicated to or from
*
* @param primaryStore the primary content store
*/
public void setPrimaryStore(ContentStore primaryStore)
{
this.primaryStore = primaryStore;
}
/**
* Set the secondary stores that this component will replicate to or from
*
* @param secondaryStores a list of stores to replicate to or from
*/
public void setSecondaryStores(List<ContentStore> secondaryStores)
{
this.secondaryStores = secondaryStores;
}
/**
* @return Returns <tt>true</tt> if the primary store supports writing
*/
public boolean isWriteSupported()
{
return primaryStore.isWriteSupported();
}
/**
* @return Returns <tt>true</tt> if the primary store supports the URL
*/
@Override
public boolean isContentUrlSupported(String contentUrl)
{
return primaryStore.isContentUrlSupported(contentUrl);
}
/**
* @return Return the primary store root location
*/
@Override
public String getRootLocation()
{
return primaryStore.getRootLocation();
}
/**
* Forwards the call directly to the first store in the list of stores.
*/
public ContentReader getReader(String contentUrl) throws ContentIOException
{
if (primaryStore == null)
{
throw new AlfrescoRuntimeException("ReplicatingContentStore not initialised");
}
// get a read lock so that we are sure that no replication is underway
readLock.lock();
try
{
// get a reader from the primary store
ContentReader primaryReader = primaryStore.getReader(contentUrl);
// give it straight back if the content is there
if (primaryReader.exists())
{
return primaryReader;
}
// the content is not in the primary reader so we have to go looking for it
for (ContentStore store : secondaryStores)
{
ContentReader reader = store.getReader(contentUrl);
if (reader.exists())
{
// found the content in a secondary store
return reader;
}
}
return primaryReader;
}
finally
{
readLock.unlock();
}
}
public boolean exists(String contentUrl)
{
if (primaryStore == null)
{
throw new AlfrescoRuntimeException("ReplicatingContentStore not initialised");
}
// get a read lock so that we are sure that no replication is underway
readLock.lock();
try
{
// Keep track of the unsupported state of the content URL - it might be a rubbish URL
boolean contentUrlSupported = false;
boolean contentUrlExists = false;
// Check the primary store
try
{
contentUrlExists = primaryStore.exists(contentUrl);
// At least the content URL was supported
contentUrlSupported = true;
}
catch (UnsupportedContentUrlException e)
{
// The store can't handle the content URL
}
if (contentUrlExists)
{
return true;
}
// the content is not in the primary store so we have to go looking for it
for (ContentStore store : secondaryStores)
{
contentUrlExists = false;
try
{
contentUrlExists = store.exists(contentUrl);
// At least the content URL was supported
contentUrlSupported = true;
}
catch (UnsupportedContentUrlException e)
{
// The store can't handle the content URL
}
if (contentUrlExists)
{
break;
}
}
// Check if the content URL was supported
if (!contentUrlSupported)
{
throw new UnsupportedContentUrlException(this, contentUrl);
}
return contentUrlExists;
}
finally
{
readLock.unlock();
}
}
public ContentWriter getWriter(ContentContext ctx)
{
// get the writer
ContentWriter writer = primaryStore.getWriter(ctx);
return writer;
}
/**
* Performs a delete on the local store and if outbound replication is on, propogates
* the delete to the other stores too.
*
* @return Returns the value returned by the delete on the primary store.
*/
public boolean delete(String contentUrl) throws ContentIOException
{
// delete on the primary store
boolean deleted = primaryStore.delete(contentUrl);
if (logger.isDebugEnabled())
{
logger.debug("Deleted content for URL: " + contentUrl);
}
return deleted;
}
}