Details
-
Type:
Sub-task
-
Status: Resolved
-
Priority:
Major
-
Resolution: Fixed
-
Affects Version/s: None
-
Fix Version/s: HDFS-5535 (Rolling upgrades)
-
Component/s: namenode
-
Labels:None
-
Hadoop Flags:Reviewed
Description
When rolling upgrade is in progress, the standby namenode may create checkpoint. The rolling upgrade information should be added to fsimage in order to support namenode restart and continue rolling upgrade.
Issue Links
- incorporates
-
HDFS-5905
Upgrade and rolling upgrade should not be allowed simultaneously
-
- Resolved
-
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
Hi Nicholas,
Since HDFS-5698 is already merged to trunk, why not merge those changes first and then make changes related to this jira using protobuf itself?
Instead of extracting checking part to checkUpgrade() can't we use nn.getFSImage().isUpgradeFinalized() directly? This will avoid checking disks on RPC calls holding writeLock.
> Since HDFS-5698 is already merged to trunk, why not merge those changes first and then make changes related to this jira using protobuf itself?
There are some conflicts with the branch. I plan to resolve them later on since I have started to work on this.
> Instead of extracting checking part to checkUpgrade() can't we use nn.getFSImage().isUpgradeFinalized() directly? This will avoid checking disks on RPC calls holding writeLock.
You brought up a good point about the write lock. However, isUpgradeFinalized() is different from checkUpgrade(), which is to check if upgrade is started. Since starting rolling upgrade is an admin command, we can safely assume that it won't be abused.
h5945_20140214.patch: allows saveNamespace since checkpoint is allowed; adds new tests.
h5945_20140216.patch: fix a bug.
Make a small change to the latest patch to use protobuf based fsimage.
+1 the protobuf change looks good.
I have committed this.
h5945_20140213.patch: adds upgrade info to fsimage and incorporates
HDFS-5905.