Suresh, Tsz-Wo,
Thanks for your comments.
Could you check a comment by Masatake on HDFS-6833?
thanks,
- Tsuyoshi
On Tue, Nov 18, 2014 at 7:14 AM, Tsz Wo Sze wrote:
> I agree that HDFS-6833 is not a blocker. It is not a problem for the blocks
> with replication >= 3 and it is not a regression
+1 (suppose using md5 is intentional.)
Verified md5 and ran a single node HA HDFS cluster for some time. Everything
worked great.
Regards,Tsz-Wo
On Monday, November 17, 2014 4:48 PM, Tsz Wo Sze
wrote:
Hi Arun,
Only md5 is provided but not mds? Is it intentional?
Tsz-Wo
On
Sorry, my vote should be non-binding.
Regards,
Yi Liu
-Original Message-
From: Liu, Yi A [mailto:yi.a@intel.com]
Sent: Tuesday, November 18, 2014 10:03 AM
To: hdfs-...@hadoop.apache.org; common-...@hadoop.apache.org;
yarn-...@hadoop.apache.org; mapreduce-dev@hadoop.apache.org
Subje
+1 (binding)
- Built from source
- Setup a security cluster with 4 nodes, and verified various fs operations.
- Verified MapReduce examples in security mode.
- Verify HDFS encryption at rest in security mode.
Regards,
Yi Liu
-Original Message-
From: Arun C Murthy [mailto:a...@hortonworks
Hi Arun,
Only md5 is provided but not mds? Is it intentional?
Tsz-Wo
On Monday, November 17, 2014 2:14 PM, Tsz Wo Sze
wrote:
I agree that HDFS-6833 is not a blocker. It is not a problem for the blocks
with replication >= 3 and it is not a regression (please correct me if I am
w
I agree that HDFS-6833 is not a blocker. It is not a problem for the blocks
with replication >= 3 and it is not a regression (please correct me if I am
wrong.)Tsz-Wo
On Monday, November 17, 2014 10:06 AM, Suresh Srinivas
wrote:
Tsuyoshi, thanks for bringing up HDFS-6833. However
+1 (binding)
Verified the signatures and hashes for both src and binary tars. Built from
the source, the binary distribution and the documentation. Started a single
node cluster and tested the following:
- Started HDFS cluster, verified the hdfs CLI commands such ls, copying
data back and forth, v
+1 (non-binding)
- Deployed single node cluster
- Verified reservation system works (YARN-1051) by making reservations and
submitting sample MR jobs against the reservations
- Verified move apps between capacity scheduler queues (YARN-2378)
Thanks,
Subru
On Mon, Nov 17, 2014 at 9:55 AM, Charles
Nada!
You can look at MergeManagerImpl.java
On Saturday, November 15, 2014 12:14 PM, Nada Saif
wrote:
Hi,
I'm interested in contributing to Hadoop ,specially MapReduce shuffle and
sort.
I'm still novice ,can you please lead me through the code ,what part of
code I should focus on?
Hi Suresh,
I did review HDFS-6833 and the patch Shinichi provided a while back. It's
not really directly related to replication factor. The problem reported
there is, DirectoryScanner tries to register with NN blocks already
scheduled to be deleted as regular blocks, due to the delay in block
remo
Tsuyoshi, thanks for bringing up HDFS-6833. However, given it is a boundary
condition (and should not cause issues when for files with replication
factor 3), we should perhaps target this into 2.6.1 and not block this
release. Thoughts?
On Mon, Nov 17, 2014 at 4:17 AM, Tsuyoshi OZAWA
wrote:
> +0
I built from the src package, started a cluster, created an encryption
zone and read/wrote data from/to it.
+1 (non-binding)
Charles
+1 (binding)
- verified signatures and digests- verified late-arriving fixes for YARN-2846
and MAPREDUCE-6156 were present
- built from source- deployed to a single-node cluster
- ran some sample MapReduce jobs
Jason
From: Arun C Murthy
To: "common-...@hadoop.apache.org" ;
"hdfs-...@hado
See https://builds.apache.org/job/Hadoop-Mapreduce-trunk/1960/
###
## LAST 60 LINES OF THE CONSOLE
###
[...truncated 30659 lines...]
Running org.apache.hadoop.mapred.Tes
Junping Du created MAPREDUCE-6164:
-
Summary: "mapreduce.reduce.shuffle.fetch.retry.timeout-ms" should
be set to 3 minutes instead of 30 seconds by default to keep insistent with
other retry timeout
Key: MAPREDUCE-6164
+0(non-binding)
HDFS-6833 is critical issue for us - could you help us to merge it into 2.6?
Thanks,
- Tsuyoshi
On Sat, Nov 15, 2014 at 1:03 PM, Hitesh Shah wrote:
> +1 (binding)
>
> Built Hadoop from source, compiled Tez against the hadoop jars pushed to
> staging repo and ran a few example T
16 matches
Mail list logo
