shard lost - solr5.3

[GOURAUD Emmanuel]

Hi there, 

I have a collection that is composed of 8 shards with a replicationFactor of 2 

i found 2 cores of the same shard in recoveryfailed status so i decided to 
restart both, 

after having doing that , i do not have any leader on that shard... and both 
cores are down 

is there a way to force a leader at startup or with the API? can fore election? 

thanks for your help 

Emmanuel 

Re: shard lost - solr5.3

[GOURAUD Emmanuel]

hi there 

replying to myself 

i have set the replica property "preferredLeader" on this shard, shut down all 
replica for this shard and started only the "preferred" one, this forced an 
election and save my "ops" night and my new year party!! 

cheers, 

Emmanuel 



De: "GOURAUD Emmanuel"  
À: solr-user@lucene.apache.org 
Envoyé: Jeudi 31 Décembre 2015 15:30:42 
Objet: shard lost - solr5.3 

Hi there, 

I have a collection that is composed of 8 shards with a replicationFactor of 2 

i found 2 cores of the same shard in recoveryfailed status so i decided to 
restart both, 

after having doing that , i do not have any leader on that shard... and both 
cores are down 

is there a way to force a leader at startup or with the API? can fore election? 

thanks for your help 

Emmanuel 

overseer status documentation

[GOURAUD Emmanuel]

Hi all, 

I'm looking for exact definitions of results returned by an "OVERSTATUS" API 
query, and the documentation talk about " various overseer APIs" 

does someone have more informations about that? 

cheers, 




Emmanuel GOURAUD 
Ingénieur Infogérance | IT Outsourcing Engineer 


egour...@jouve.fr 


Tél. : +33(0) 2 43 08 25 54 



1, rue du docteur Sauvé, 53100 Mayenne, France 
www.jouve.com   
Avant d'imprimer cet e-mail, réfléchissez à l'impact sur l'environnement. 
Merci. Before you print this e-mail, think about the impact on the environment. 
Thank you. 
Le présent mail ainsi que toutes les informations qu'il contient ne peuvent en 
aucun cas être considérés comme un engagement juridique de quelque nature que 
ce soit de JOUVE. Tout accord devra être formulé par écrit papier ultérieur 
signé par un représentant légal de JOUVE. Par ailleurs, si vous recevez ce mail 
par erreur, merci de nous le signaler et de le détruire ainsi que l'intégralité 
du document qui pourrait y être joint. 
The present email and all information included therein do not constitute a 
legal agreement accorded by Jouve. All legal agreements must be formulated in 
writing on paper by a legal representative of JOUVE. If you have received this 
email by mistake, please inform us of that fact and destroy the email and any 
documents it might contain. Thank you for your cooperation. 

restoring an index from restore command

[GOURAUD Emmanuel]

Hi there, 

I'm currently trying to restore a core from a snapshot, and it does not work as 
excepted (in my case). 

My context: 
I'm using solr 5.3.0 
i used named snapshot and specific location to ensure collection integrity a a 
time in a single directory 

below, i will describe backup/restore for a specific core 

backup command: 
http://10.0.2.15:8080/solr/test_shard2_replica1/replication?wt=json&command=backup&location=%2Fhome%2Fsolr%2Ftest%2F20151128-073603009560&name=test_shard2_replica1
 

it has created the snapshot with all data: 
[vagrant@centos7 solr-5.3.0]$ ls -lh 
/home/solr/test/20151128-073603009560/snapshot.test_shard1_replica2/ 
total 32K 
-rw-r--r-- 1 solr solr 346 Nov 28 07:36 _5.fdt 
-rw-r--r-- 1 solr solr 84 Nov 28 07:36 _5.fdx 
-rw-r--r-- 1 solr solr 431 Nov 28 07:36 _5.fnm 
-rw-r--r-- 1 solr solr 110 Nov 28 07:36 _5_Lucene50_0.doc 
-rw-r--r-- 1 solr solr 697 Nov 28 07:36 _5_Lucene50_0.tim 
-rw-r--r-- 1 solr solr 171 Nov 28 07:36 _5_Lucene50_0.tip 
-rw-r--r-- 1 solr solr 432 Nov 28 07:36 _5.si 
-rw-r--r-- 1 solr solr 165 Nov 28 07:36 segments_7 

i flush all data from the same core and restart the solr5 instance and then try 
to restore from the previous snapshot: 
[vagrant@centos7 solr-5.3.0]$ curl 
"http://10.20.1.16:8082/solr/test_shard1_replica2/replication?wt=json&command=restore&location=%2Fhome%2Fsolr%2Ftest%2F20151128-073603009560&name=test_shard1_replica2";
 
{"responseHeader":{"status":0,"QTime":53},"status":"OK"} 

all seems to be OK, let's verify 

[vagrant@centos7 solr-5.3.0]$ curl 
"http://10.20.1.16:8082/solr/test_shard1_replica2/replication?wt=json&command=restorestatus"{"responseHeader":{"status":0,"QTime":2},"restorestatus":{"snapshotName":"snapshot.test_shard1_replica2","status":"success"}}
 

great!! what about logs: 

[...skipped data...] 
2015-11-28 07:42:39.128 INFO 
(restoreExecutor-8-thread-1-processing-x:test_shard1_replica2 r:core_node4 
test_shard1_replica2 n:10.20.1.16:8082_solr snapshot.test_shard1_replic 
a2 s:shard1 c:test //home//solr//test//20151128-073603009560) [c:test s:shard1 
r:core_node4 x:test_shard1_replica2] o.a.s.h.RestoreCore Copying file _5.fdt to 
restore directory 

2015-11-28 07:42:39.129 INFO 
(restoreExecutor-8-thread-1-processing-x:test_shard1_replica2 r:core_node4 
test_shard1_replica2 n:10.20.1.16:8082_solr snapshot.test_shard1_replic 
a2 s:shard1 c:test //home//solr//test//20151128-073603009560) [c:test s:shard1 
r:core_node4 x:test_shard1_replica2] o.a.s.h.RestoreCore Copying file 
_5_Lucene50_0.tip to restor 
e directory 
2015-11-28 07:42:39.129 INFO 
(restoreExecutor-8-thread-1-processing-x:test_shard1_replica2 r:core_node4 
test_shard1_replica2 n:10.20.1.16:8082_solr snapshot.test_shard1_replic 
a2 s:shard1 c:test //home//solr//test//20151128-073603009560) [c:test s:shard1 
r:core_node4 x:test_shard1_replica2] o.a.s.h.RestoreCore Copying file _5.si to 
restore directory 
2015-11-28 07:42:39.130 INFO 
(restoreExecutor-8-thread-1-processing-x:test_shard1_replica2 r:core_node4 
test_shard1_replica2 n:10.20.1.16:8082_solr snapshot.test_shard1_replic 
a2 s:shard1 c:test //home//solr//test//20151128-073603009560) [c:test s:shard1 
r:core_node4 x:test_shard1_replica2] o.a.s.h.RestoreCore Copying file 
_5_Lucene50_0.doc to restor 
e directory 
2015-11-28 07:42:39.132 INFO 
(restoreExecutor-8-thread-1-processing-x:test_shard1_replica2 r:core_node4 
test_shard1_replica2 n:10.20.1.16:8082_solr snapshot.test_shard1_replic 
a2 s:shard1 c:test //home//solr//test//20151128-073603009560) [c:test s:shard1 
r:core_node4 x:test_shard1_replica2] o.a.s.h.RestoreCore Copying file 
_5_Lucene50_0.tim to restor 
e directory 
2015-11-28 07:42:39.134 INFO 
(restoreExecutor-8-thread-1-processing-x:test_shard1_replica2 r:core_node4 
test_shard1_replica2 n:10.20.1.16:8082_solr snapshot.test_shard1_replic 
a2 s:shard1 c:test //home//solr//test//20151128-073603009560) [c:test s:shard1 
r:core_node4 x:test_shard1_replica2] o.a.s.h.RestoreCore Copying file _5.fdx to 
restore directory 
[...skipped data...] 

Wow , i love solr ;) 

now, let's have a look at data directory: 

[vagrant@centos7 solr-5.3.0]$ ls -lh 
/usr/local/third/solr-5.3.0/server/solr/test_shard1_replica2/data/* 
-rw-r--r-- 1 solr solr 92 Nov 28 07:42 
/usr/local/third/solr-5.3.0/server/solr/test_shard1_replica2/data/index.properties
 

/usr/local/third/solr-5.3.0/server/solr/test_shard1_replica2/data/restore.snapshot.test_shard1_replica2:
 
total 0 
-rw-r--r-- 1 solr solr 0 Nov 28 07:42 write.lock 

/usr/local/third/solr-5.3.0/server/solr/test_shard1_replica2/data/tlog: 
total 0 

Hum, no data, what did i miss? 

in order to workaround, i can restore from filesystem, but this behavior is 
making me mad :) 

thanks for your help ;) 

Re: Solrcloud with Zookeeper in production

[GOURAUD Emmanuel]

Hi, 

bootstrap_confdir is used to send config files into zookeeper at the start of 
the solr instance, however you should notice that if you modify config files 
into zookeeper you'll need t restart all cores to load changes. Moreover, you 
must restart the designed core (with bootstrap_cmd) first to ensure config 
files are updated before restarting the others. 

You should use zkcli.sh to import (upconfig) your config files into zookeeper 
and then restart associated cores (whatever the order). 
https://cwiki.apache.org/confluence/display/solr/Using+ZooKeeper+to+Manage+Configuration+Files
 




De: "Mugeesh Husain"  
À: solr-user@lucene.apache.org 
Envoyé: Dimanche 29 Novembre 2015 09:20:21 
Objet: Re: Solrcloud with Zookeeper in production 

Hi, 

I was following this article 
http://jayant7k.blogspot.in/2013/06/step-by-step-setting-up-solr-cloud.html 
for solr.5.3 +zk-3.4.6 version. 

in the above linke, there is an step for configuration files 
root@solr1$ java -DzkHost=solr1:2181,solr2:2181 
-Dbootstrap_confdir=solr/collection1/conf/ -DnumShards=2 - 

I noticed "-Dbootstrap_confdir=solr/collection1/conf/" , i guess this is the 
location of configuration files, 

could you give such type of step for solr 5.3. 

Actually I have tried to many method unable to configure configuration files 
for core. 
i am looking for that step. 




-- 
View this message in context: 
http://lucene.472066.n3.nabble.com/Solrcloud-with-Zookeeper-in-production-tp4242162p4242596.html
 
Sent from the Solr - User mailing list archive at Nabble.com.