Hard disk crashes in elastic search


(soumya) #1

In elastic search if a node crashes then data can be recovered from it
replicas.
But what happens if a single node's hard disk crashes and data is
lost.
Is there any mechanism to recover data then ?

Thanks in advance for any help

Soumya


(Shay Banon) #2

You mean you run a single node cluster, and the data is lost? No, in this
case, there is no way to recover the data.

On Fri, Aug 12, 2011 at 3:15 PM, Soumya MR SoumyaMR@algotree.com wrote:

In elastic search if a node crashes then data can be recovered from it
replicas.
But what happens if a single node's hard disk crashes and data is
lost.
Is there any mechanism to recover data then ?

Thanks in advance for any help

Soumya


(James Cook) #3

Except with a gateway writing to another server.


(CHI WAI HO) #4

How about if there is data file corruption in a cluster with
replicated nodes, will the error be replicated to other healthy node?

Thanks a lot.

On 8月14日, 上午4時04分, James Cook jc...@tracermedia.com wrote:

Except with a gateway writing to another server.


(Shay Banon) #5

Because elasticsearch does push replication, other shards index the data as
well, so a corruption will not propagate to other nodes.

On Sun, Aug 14, 2011 at 4:24 AM, CHI WAI HO zeroho@gmail.com wrote:

How about if there is data file corruption in a cluster with
replicated nodes, will the error be replicated to other healthy node?

Thanks a lot.

On 8月14日, 上午4時04分, James Cook jc...@tracermedia.com wrote:

Except with a gateway writing to another server.


(system) #6