hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Arun Ramakrishnan <aramakrish...@languageweaver.com>
Subject RE: rebalancing replication help
Date Thu, 08 Jul 2010 18:11:13 GMT
Thanks Edward.

The setting to configure how aggressively replication is performed is dfs.balance.bandwidthPerSec
? I know it's used by the balancer. But, I can't find any other parameter that makes sense
in this context.

Thanks
Arun

-----Original Message-----
From: Edward Capriolo [mailto:edlinuxguru@gmail.com] 
Sent: Wednesday, July 07, 2010 6:58 PM
To: common-user@hadoop.apache.org
Cc: general@hadoop.apache.org
Subject: Re: rebalancing replication help

On Wed, Jul 7, 2010 at 9:18 PM, Arun Ramakrishnan
<aramakrishnan@languageweaver.com> wrote:
> Looks like there is not much activity in the hdfs-user list. So, am reposting it in the
general list.
>
> Hi guys.
>  I have a few related questions. I am going to layout the steps I have taken. Please
comment on what I can do better.
>
>  I was trying to to add 5 nodes to my existing 10 node cluster and also increase the
replication factor from 2 to 3.
> I thought I don't have to run the balancer cause it would most likely put the new replicas
into the new nodes.
>
> There are about 500k blocks.
> I wanted to get it all stabilized(replication and balancing) within 24 hours. Its more
than 24 hours now and fsck reports 30% under replication. Is there a way to force hdfs to
use balance/replicate more aggressively.
>
> It would be great if someone explained what/when things happen to blocks in the context
of
>
> 1)      Rebalancing
>
> 2)      -setrep
>
> 3)      Restarting cluster with a higher/lower replication factor.
>
> A few questions and a few issues here.
>
> 1)      When you restart the cluster with a higher than previous replication value.
Does it also apply to existing blocks or only to new blocks being created ?
>
> 2)      Does the balancer take into account under replication of blocks or does it
blindly start moving existing blocks to reach threshold ?
>
>
> A very specific problem .  I am having this strange problem where the -setrep hangs
on one particular block for hours. Is this because its corrupt ?. But, fsck said its healthy.
>
>
> Thanks
> Arun
>
>

> 2)      -setrep
This will change the replication factor of an existing file (in the
background it should start replicating)

> 2) Does the balancer take into account under replication of blocks or does it blindly
start moving existing blocks to reach threshold ?

Files most under replication should be prioritized.

> 3)      Restarting cluster with a higher/lower replication factor.
This only affects new files that are created. Where the client has not
specified a value

> A very specific problem .  I am having this strange problem where the -setrep hangs on
one particular block for hours. Is this because its corrupt ?. But, fsck said its healthy.
Not sure

> Its more than 24 hours now and fsck reports 30% under
There is a configuration setting for maximum replication bandwidth.
You might have to tune that.

Mime
View raw message