Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: pass (nike.apache.org: domain of nidmgg@gmail.com designates
 209.85.128.176 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CAEbjYSDm4WYOL2s9bqnpCDWoda+cinVo9v3MkOV_G7vM_PRH_w@mail.gmail.com>
References: 
 <CAEbjYSDm4WYOL2s9bqnpCDWoda+cinVo9v3MkOV_G7vM_PRH_w@mail.gmail.com>
Date: Mon, 28 Oct 2013 14:45:03 -0700
Message-ID: 
 <CAOEq2C5OAn6GXUzsWLFz6WgOpB7sbC-PCmhLgSL2ncvvcr=TRA@mail.gmail.com>
Subject: Re: HBase Replication Questions
From: Demai Ni <nidmgg@gmail.com>
To: user@hbase.apache.org
Content-Type: multipart/alternative; boundary=20cf3079b5f236e0c204e9d403b8

--20cf3079b5f236e0c204e9d403b8
Content-Type: text/plain; charset=ISO-8859-1

1)  you can use replication metrics info such as:ageOfLastShippedOp,
timeStampsOfLastShippedOp, sizeOfLogQueue ageOfLastAppliedOp, and
timeStampsOfLastAppliedOp to figure out the lag.
2) with my best knowledge, the replication won't be together with snapshot.
That is the peers won't know there is a restore occurred on primary. So a
user must take care it manually.   the VerifyReplication tool may be
helpful to identify which data (boundary by timestamps) is mismatch, so the
copyTable doesn't have to copy the whole table

Demai

Demai


On Mon, Oct 28, 2013 at 12:07 PM, hdev ml <hdevml@gmail.com> wrote:

> Hello all,
>
> I have a couple of questions for HBase.
>
> 1. Is there any way to find out by how much time the replication is lagging
> on the peer cluster or even primary cluster is ok.
> 2. Say replication is going on for some time for a table and then you
> restore a snapshot on the Primary. Would replication catch up with the
> restore snapshot? Or do I have to treat that as replicating from scratch,
> meaning
>
>    1. disable peer
>    2. restore snapshot on primary
>    3. copyTable upto disable timestamp
>    4. enable peer
>
>
> Any help is appreciated.
>
> Thanks
>
> Harshad
>

--20cf3079b5f236e0c204e9d403b8--