[
https://issues.apache.org/jira/browse/DL-37?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15477272#comment-15477272
]
ASF GitHub Bot commented on DL-37:
----------------------------------
Github user leighst commented on a diff in the pull request:
https://github.com/apache/incubator-distributedlog/pull/15#discussion_r78195828
--- Diff:
distributedlog-core/src/main/java/com/twitter/distributedlog/util/OffsetSequencer.java
---
@@ -0,0 +1,51 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package com.twitter.distributedlog.util;
+
+import com.twitter.distributedlog.LogRecord;
+
+/**
+ * Offset based sequencer. It generated non-decreasing transaction id
using offsets.
+ * It isn't thread-safe. The caller takes the responsibility on
synchronization.
+ */
+public class OffsetSequencer implements Sequencer {
+
+ private long lastOffset = -1L;
+
+ @Override
+ public void setLastId(long lastId) {
+ // NOTE: it is a bit tricky here. when a log stream is recovered,
we can only get the last transaction id,
+ // but not the length of last record. so we don't know how
many `bytes` to advance. so just to advance
+ // one here.
+ this.lastOffset = lastId + 1;
+ }
+
+ @Override
+ public long nextId() {
+ return lastOffset;
+ }
+
+ @Override
+ public void advance(LogRecord record) {
+ // skip the control records (they are invisible to the users)
+ if (record.isControl()) {
+ return;
+ }
+ this.lastOffset += record.getPayload().length;
--- End diff --
It is a little tricky to support this reliably in write proxy (depends what
guarantees you need) -
If some writes fail (ex. due to LogRecordTooLongException) you will have
gaps in your offsets - is this ok ?
> Add OffsetSequencer
> -------------------
>
> Key: DL-37
> URL: https://issues.apache.org/jira/browse/DL-37
> Project: DistributedLog
> Issue Type: Improvement
> Components: distributedlog-core, distributedlog-service
> Reporter: Khurrum Nasim
> Assignee: Khurrum Nasim
> Fix For: 0.4.0
>
>
> In order to support kafka like offset, we'd like to introduce an offset
> sequencer.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
