Upgrade migration path?

IvanBrusic · December 14, 2010, 10:52pm

Due to a change in the API, I decided to upgrade from 0.12.1 to
0.13.1. I copied the entire work directory plus my custom mappings
from config into the new install directory.

Startup seems fine, but my client is showing issues:

17:49:45.846 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [0] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@2e56c4eb, error [false], resetting
17:49:45,858 ERROR ~ Missing ad hosts file: data/adhosts/adhosts.dat
17:49:46.725 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [1] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@3267b4ba, error [false], resetting
17:49:47.728 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [2] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@1cbcfd, error [false], resetting
17:49:48.731 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [3] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@696103e5, error [false], resetting
17:49:49.732 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [4] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@6046e56f, error [false], resetting
...

gist.github.com

https://gist.github.com/ibrusic/741273

gistfile1.txt

[2010-12-14 17:46:09,463][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: initializing ...
[2010-12-14 17:46:09,466][INFO ][plugins                  ] [Stonewall] loaded []
[2010-12-14 17:46:10,747][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: initialized
[2010-12-14 17:46:10,747][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: starting ...
[2010-12-14 17:46:10,805][INFO ][transport                ] [Stonewall] bound_address {inet[/0.0.0.0:9300]}, publish_address {inet[/192.168.1.103:9300]}
[2010-12-14 17:46:13,826][INFO ][cluster.service          ] [Stonewall] new_master [Stonewall][JsZdjwSISaqoMkZJH5G-Jw][inet[/192.168.1.103:9300]], reason: zen-disco-join (elected_as_master)
[2010-12-14 17:46:13,843][INFO ][discovery                ] [Stonewall] elasticsearch/JsZdjwSISaqoMkZJH5G-Jw
[2010-12-14 17:46:13,868][INFO ][http                     ] [Stonewall] bound_address {inet[/0.0.0.0:9200]}, publish_address {inet[/192.168.1.103:9200]}
[2010-12-14 17:46:13,868][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: started

Java code is simple:

String ELASTIC_SEARCH_HOST = "localhost"
int ELASTIC_SEARCH_PORT = 9300

TransportClient client = new

TransportClient().addTransportAddress(new
InetSocketTransportAddress(ELASTIC_SEARCH_HOST, ELASTIC_SEARCH_PORT))

Is copying the work directory acceptable when moving from 0.12.1 to
0.13.1?

Cheers,

Ivan

kimchy · December 15, 2010, 5:41pm

Have you looked at the upgrade notes (work became data)? Also, are you sure
you are running the same version of ES on the client and server side?

On Wednesday, December 15, 2010 at 12:52 AM, Ivan Brusic wrote:

Due to a change in the API, I decided to upgrade from 0.12.1 to
0.13.1. I copied the entire work directory plus my custom mappings
from config into the new install directory.

Startup seems fine, but my client is showing issues:

17:49:45.846 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [0] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@2e56c4eb, error [false], resetting
17:49:45,858 ERROR ~ Missing ad hosts file: data/adhosts/adhosts.dat
17:49:46.725 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [1] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@3267b4ba, error [false], resetting
17:49:47.728 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [2] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@1cbcfd, error [false], resetting
17:49:48.731 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [3] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@696103e5, error [false], resetting
17:49:49.732 [New I/O client worker #1-1] WARN transport.netty -
[Warlock] Message not fully read (response) for [4] handler
org.elasticsearch.client.transport.TransportClientNodesService
$ScheduledNodesSampler$1$1@6046e56f, error [false], resetting
...

gist.github.com

https://gist.github.com/ibrusic/741273

gistfile1.txt

[2010-12-14 17:46:09,463][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: initializing ...
[2010-12-14 17:46:09,466][INFO ][plugins                  ] [Stonewall] loaded []
[2010-12-14 17:46:10,747][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: initialized
[2010-12-14 17:46:10,747][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: starting ...
[2010-12-14 17:46:10,805][INFO ][transport                ] [Stonewall] bound_address {inet[/0.0.0.0:9300]}, publish_address {inet[/192.168.1.103:9300]}
[2010-12-14 17:46:13,826][INFO ][cluster.service          ] [Stonewall] new_master [Stonewall][JsZdjwSISaqoMkZJH5G-Jw][inet[/192.168.1.103:9300]], reason: zen-disco-join (elected_as_master)
[2010-12-14 17:46:13,843][INFO ][discovery                ] [Stonewall] elasticsearch/JsZdjwSISaqoMkZJH5G-Jw
[2010-12-14 17:46:13,868][INFO ][http                     ] [Stonewall] bound_address {inet[/0.0.0.0:9200]}, publish_address {inet[/192.168.1.103:9200]}
[2010-12-14 17:46:13,868][INFO ][node                     ] [Stonewall] {elasticsearch/0.13.1}[45456]: started

Java code is simple:

String ELASTIC_SEARCH_HOST = "localhost"
int ELASTIC_SEARCH_PORT = 9300

TransportClient client = new
TransportClient().addTransportAddress(new
InetSocketTransportAddress(ELASTIC_SEARCH_HOST, ELASTIC_SEARCH_PORT))

Is copying the work directory acceptable when moving from 0.12.1 to
0.13.1?

Cheers,

Ivan