Hi,
a big thank you to Jeon Woosung - we just upgraded our cloud to 4.10.2.
One correction: we had to use
/collections/{collection}/leader_initiated_recovery/shard1/node5, where
"node5" had to be replaced with the place the down node showed up in the
solr cloud dashboard. Also no tomcat restart was neccessary - even
contra productive, since state changes may overwrite the just-fixed enty.
Best regards
Thomas
Am 13.11.2014 um 05:47 schrieb Jeon Woosung:
you can migrate zookeeper data manually.
1. connect zookeeper.
- zkCli.sh -server host:port
2. check old data
- get /collections/"your collection
name"/leader_initiated_recovery/"your shard name"
================================================================================
[zk: localhost:3181(CONNECTED) 25] get
/collections/collection1/leader_initiated_recovery/shard1
*down*
cZxid = 0xe4
ctime = Thu Nov 13 13:38:53 KST 2014
mZxid = 0xe4
mtime = Thu Nov 13 13:38:53 KST 2014
pZxid = 0xe4
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 4
numChildren = 0
================================================================================
i guess that there is only single word which is "down"
3. delete the data.
- remove /collections/"your collection
name"/leader_initiated_recovery/"your shard name"
4. create new data.
- create /collections/"your collection
name"/leader_initiated_recovery/"your shard name" {state:down}
5. restart the server.
On Thu, Nov 13, 2014 at 7:42 AM, Anshum Gupta <ans...@anshumgupta.net>
wrote:
Considering the impact, I think we should put this out as an announcement
on the 'news' section of the website warning people about this.
On Wed, Nov 12, 2014 at 12:33 PM, Shalin Shekhar Mangar <
shalinman...@gmail.com> wrote:
I opened https://issues.apache.org/jira/browse/SOLR-6732
On Wed, Nov 12, 2014 at 12:29 PM, Shalin Shekhar Mangar <
shalinman...@gmail.com> wrote:
Hi Thomas,
You're right, there's a back-compat break here. I'll open an issue.
On Wed, Nov 12, 2014 at 9:37 AM, Thomas Lamy <t.l...@cytainment.de>
wrote:
Am 12.11.2014 um 15:29 schrieb Thomas Lamy:
Hi there!
As we got bitten by https://issues.apache.org/jira/browse/SOLR-6530
on
a regular basis, we started upgrading our 7 mode cloud from 4.10.1 to
4.10.2.
The first node upgrade worked like a charm.
After upgrading the second node, two cores no longer come up and we
get
the following error:
ERROR - 2014-11-12 15:17:34.226;
org.apache.solr.cloud.RecoveryStrategy;
Recovery failed - trying again... (16) core=cams_shard1_replica4
ERROR - 2014-11-12 15:17:34.230;
org.apache.solr.common.SolrException;
Error while trying to recover. core=onlinelist_shard1_
replica7rg.noggit.JSONParser$ParseException: JSON Parse Error:
char=d,position=0 BEFORE='d' AFTER='own'
at org.noggit.JSONParser.err(JSONParser.java:223)
at org.noggit.JSONParser.next(JSONParser.java:622)
at org.noggit.JSONParser.nextEvent(JSONParser.java:663)
at org.noggit.ObjectBuilder.<init>(ObjectBuilder.java:44)
at org.noggit.ObjectBuilder.getVal(ObjectBuilder.java:37)
at org.apache.solr.common.cloud.ZkStateReader.fromJSON(
ZkStateReader.java:129)
at
org.apache.solr.cloud.ZkController.getLeaderInitiatedRecoveryStat
eObject(ZkController.java:1925)
at
org.apache.solr.cloud.ZkController.getLeaderInitiatedRecoveryStat
e(ZkController.java:1890)
at org.apache.solr.cloud.ZkController.publish(
ZkController.java:1071)
at org.apache.solr.cloud.ZkController.publish(
ZkController.java:1041)
at org.apache.solr.cloud.ZkController.publish(
ZkController.java:1037)
at org.apache.solr.cloud.RecoveryStrategy.doRecovery(
RecoveryStrategy.java:355)
at org.apache.solr.cloud.RecoveryStrategy.run(
RecoveryStrategy.java:235)
Any hint on how to solve this? Google didn't reveal anything
useful...
Kind regards
Thomas
Just switched to INFO loglevel:
INFO - 2014-11-12 15:30:31.563;
org.apache.solr.cloud.RecoveryStrategy;
Publishing state of core onlinelist_shard1_replica7 as recovering,
leader
is http://solr-bc1-blade2:8080/solr/onlinelist_shard1_replica2/ and I
am
http://solr-bc1-blade3:8080/solr/onlinelist_shard1_replica7/
INFO - 2014-11-12 15:30:31.563;
org.apache.solr.cloud.RecoveryStrategy;
Publishing state of core cams_shard1_replica4 as recovering, leader is
http://solr-bc1-blade2:8080/solr/cams_shard1_replica2/ and I am
http://solr-bc1-blade3:8080/solr/cams_shard1_replica4/
INFO - 2014-11-12 15:30:31.563; org.apache.solr.cloud.ZkController;
publishing core=onlinelist_shard1_replica7 state=recovering
collection=onlinelist
INFO - 2014-11-12 15:30:31.563; org.apache.solr.cloud.ZkController;
publishing core=cams_shard1_replica4 state=recovering collection=cams
ERROR - 2014-11-12 15:30:31.564; org.apache.solr.common.SolrException;
Error while trying to recover. core=cams_shard1_replica4rg.
noggit.JSONParser$ParseException: JSON Parse Error: char=d,position=0
BEFORE='d' AFTER='own'
ERROR - 2014-11-12 15:30:31.564; org.apache.solr.common.SolrException;
Error while trying to recover. core=onlinelist_shard1_
replica7rg.noggit.JSONParser$ParseException: JSON Parse Error:
char=d,position=0 BEFORE='d' AFTER='own'
ERROR - 2014-11-12 15:30:31.564;
org.apache.solr.cloud.RecoveryStrategy;
Recovery failed - trying again... (5) core=cams_shard1_replica4
ERROR - 2014-11-12 15:30:31.564;
org.apache.solr.cloud.RecoveryStrategy;
Recovery failed - trying again... (5) core=onlinelist_shard1_replica7
INFO - 2014-11-12 15:30:31.564;
org.apache.solr.cloud.RecoveryStrategy;
Wait 60.0 seconds before trying to recover again (6)
INFO - 2014-11-12 15:30:31.564;
org.apache.solr.cloud.RecoveryStrategy;
Wait 60.0 seconds before trying to recover again (6)
The leader for both collections (solr-bc1-blade2) is still on 4.10.1.
As no special instructions were given in the release notes and it's a
minor upgrade, we thought there should be no BC issues and planned to
upgrade one node after the other.
Did that provide more insight?
--
Thomas Lamy
Cytainment AG & Co KG
Nordkanalstrasse 52
20097 Hamburg
Tel.: +49 (40) 23 706-747
Fax: +49 (40) 23 706-139
Sitz und Registergericht Hamburg
HRA 98121
HRB 86068
Ust-ID: DE213009476
--
Regards,
Shalin Shekhar Mangar.
--
Regards,
Shalin Shekhar Mangar.
--
Anshum Gupta
http://about.me/anshumgupta
--
Thomas Lamy
Cytainment AG & Co KG
Nordkanalstrasse 52
20097 Hamburg
Tel.: +49 (40) 23 706-747
Fax: +49 (40) 23 706-139
Sitz und Registergericht Hamburg
HRA 98121
HRB 86068
Ust-ID: DE213009476
