liangyepianzhou commented on code in PR #15239:
URL: https://github.com/apache/pulsar/pull/15239#discussion_r855938656
##########
site2/docs/tiered-storage-filesystem.md:
##########
@@ -520,4 +520,40 @@ Execute the following commands in the repository where you
download Pulsar tarba
And the **Capacity Used** is changed from 4 KB to 116.46 KB.
- 
\ No newline at end of file
+ 
+
+## Read data from fileSystem
+
+This section provides detailed instructions about how to read data out as
Ledger Entry in the file system.
+
+* The data was offloaded as `MapFile` to the following path:
+ ```properties
+ path = storageBasePath + "/" + managedLedgerName + "/" + ledgerId + "-" +
uuid.toString();
+ ```
+ 1. storageBasePath is the value of `hadoop.tmp.dir`, configured in
`broker.conf` and `filesystem_offload_core_site.xml`
+ 2. managedLedgerName is the name of the persistentTopic manager Ledger
+ ```shell
+ managedLedgerName of persistent://public/default/topics-name is
public/default/persistent/topics-name.
+ ```
+ Considering the iteration of versions, you can use the following method to
get the managedLedgerName:
+ ```shell
+ String managedLedgerName =
TopicName.get("persistent://public/default/topics-name").getPersistenceNamingEncoding();
+ ```
+
+* Create a reader to read `MapFile` according to the above path and the
`configuration` of the file system
+ ```shell
+ MapFile.Reader reader = new MapFile.Reader(new Path(dataFilePath),
configuration);
+ ```
+* Read data as `LedgerEntry` from FileSystem.
+ ```java
+ LongWritable key = new LongWritable();
+ BytesWritable value = new BytesWritable();
+ key.set(nextExpectedId - 1);
+ reader.seek(key);
+ reader.next(key, value);
+ int length = value.getLength();
+ long entryId = key.get();
+ ByteBuf buf = PooledByteBufAllocator.DEFAULT.buffer(length, length);
+ buf.writeBytes(value.copyBytes());
+ LedgerEntryImpl ledgerEntry = LedgerEntryImpl.create(ledgerId, entryId,
length, buf);
Review Comment:
Great advice. I have added the deserialization steps, please review again
when you have time.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
