pretty sure that's not
even necessary to use ES with Spark it's just a convenience if you already
have a YARN cluster and want to provision and control elasticsearch within
that environment.
Kimbro Staken
On Mon, Apr 27, 2015 at 11:03 AM, Garrett Gottlieb
garrett.gottl...@gmail.com wrote:
Hi
, Kimbro Staken kst...@kstaken.com
wrote:
Hello Brian,
Many things will affect the rate of ingest, the biggest one is making
sure the load gets spread around. But are you sure ES is what's
bottlenecking here? With only 5 shards you're only using half your cluster
but I'm willing to bet your
and
fielddata can eat heap in ways that will make your head spin.
Kimbro Staken
On Wed, Apr 22, 2015 at 1:14 AM, fdevilla...@synthesio.com wrote:
Hi list,
I've been using ES in production since 0.17.6 with clusters up to 64
virtual machines and 20T data (including 3 replica). We're now thinking
current numbers
don't show that.
Kimbro Staken
On Wed, Apr 22, 2015 at 4:16 PM, bparki...@maprtech.com wrote:
We are running a 10-node Elasticsearch 1.4.2 cluster, and getting cluster
wide throughput of 18161 docs/sec, or about 18MB/sec. We'd like to improve
this as much as we can
. An individual node can
(and should) hold dozens of shards. Larger shard sizes will work too but
when a node crashes recovery of a larger number of 50GB shards will be much
faster than a smaller number of 200GB shards, especially in a large cluster.
Kimbro Staken
On Wed, Apr 22, 2015 at 6:04 PM, Jack Park
Yes, I saw the same issue yesterday on a test system. For me it started
after the node crashed and rebooted. It looks like it's trying to use the
IPv6 address to connect. I didn't really dig too far for a real fix since
this was a test system but setting network.bind_host 0.0.0.0 got marvel
How many shards to use is a complicated question and depends on the
specific use case. For testing in this scenario though, it's likely that
just matching the number of nodes you have would be a good choice. Then you
will have 1 primary shard for each index on each node.
That said it also looks
Your second example is what elasticsearch will do by default. It will never
allocate a primary and replica for the same shard on the same node. In that
example if one of the nodes were to go down both primaries would move to
the remaining node and the replicas would be unallocated and the cluster
Yes you have something very wrong. That is showing you have a huge number
of shards and the cluster is obviously struggling to allocate all of them.
You said you have 9 nodes and 1 replica but you didn't specify how many
shards per index?
Kimbro
On Thu, Jan 22, 2015 at 11:45 AM, Sam Flint
I've experienced what you're describing. I called it a shard relocation
storm and it's really tough to get under control. I opened a ticket on the
issue and a fix was supposedly included in 1.4.2. What version are you
running?
If you want to truly manually manage this situation you could set
with more disk space to
avoid this situations.
By any chance do you have the link to the issue?
2015-01-15 13:26 GMT-03:00 Kimbro Staken ksta...@kstaken.com:
I've experienced what you're describing. I called it a shard relocation
storm and it's really tough to get under control. I opened
Looks like you have an extra opening brace under properties.
On Tue, Dec 9, 2014 at 1:36 PM, Mohit Anchlia mohitanch...@gmail.com
wrote:
I am trying to change the mapping but I think I am not doing it right.
Documents are not clear either:
My document looks like this:
This is caused by elasticsearch trying to load fielddata. Fielddata is used
for sorting and faceting/aggregations. When a query has a sort parameter
the node will try to load the fielddata for that field for all documents in
the shard, not just those included in the query result. The breaker is
13 matches
Mail list logo