Mongo::Error::OperationFailure: Cursor not found

Lately, I’ve been running into this error while running my nightly automation scripts. From my experience in resolving nagging errors, this one too was another of those annoying/inconsistent errors which did not have any concrete solutions on the internet.  This post is for the benefit of those fortunate people(unlike me) who will encounter this error in the future.

Small Background on the task:I had to query all data which was in my MongoDB server one-by-one and compare it on real-time with my api responses. I am using the ‘mongo’ ruby driver gem to interact with the db.The total data in the db was ~3.5 lac records but while running the script – at around the ~350 iteration mark I was getting this error :-

Mongo::Error::OperationFailure:
  Cursor not found, cursor id: 79727049273 (43)
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/operation/result.rb:256:in `validate!'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/operation/executable.rb:37:in `block in execute'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/server/connection_pool.rb:107:in `with_connection'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/server.rb:242:in `with_connection'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/operation/executable.rb:35:in `execute'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/cursor.rb:188:in `block in get_more'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/retryable.rb:51:in `read_with_retry'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/cursor.rb:187:in `get_more'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/cursor.rb:113:in `each'
/home/qaserver/.rvm/gems/ruby-2.3.0/gems/mongo-2.4.0/lib/mongo/collection/view/iterable.rb:44:in `each'
./spec/all_usecases_spec/rovi_ott_validation_spec/rovi_ott_links_validation_for_all_programs_spec.rb:66:in `block (2 levels) in '

div>

254      # @since 2.0.0
255      def validate!
256        !successful? ? raise(Error::OperationFailure.new(parser.message)) : self
257      end
258
259# Install the coderay gem to get syntax highlightingpre>

The crazy part was this error was not at all consistent but would happen at times. I would overlook this by re-running  my scripts.One day suddenly out of nowhere this error became almost 100% consistent! On priority, I had to find a solution for it.

After doing some amount of reading, I realised that this error is to do with the cursor which gets created while querying the db. So what happens is MongoDB returns a cursor when the query happens. In my case as my query is one which ‘finds all’ I do not fully know if multiple cursors are returned for each sub-query or a single cursor is returned which loops through the whole db. I need some more clarification on that.

But what I understand is that MongoDB closes all cursors that have been inactive for 10 minutes.It has something called a cursor timeout to do the same. So maybe one such cursor created was getting inactive after a particular time.

On more exploring I understood that there is a way to disable this cursor timeout. The hard part was to find the key word for this cursor timeout for the ruby driver which I was using, in my case the ruby driver ‘mongo’. Going through multiple stackoverflows which gave some incorrect solutions like use ‘:timeout => false’ I had to struggle my way to find this answer.

After going through the Mongo Ruby Driver documentation(which has a very confusing sequence) thouroghly, I found my answer!

There is an option while querying called ‘no_cursor_timeout’ which must be used to disable this cursor timeout. Here’s how you implement it :-

coll.find({:date => { ‘$eq’ => Date.today }}).no_cursor_timeout.each do |doc|

          ########## Code goes in here ###########

end

sudo /etc/init.d/jenkins start not working

I had installed jenkins and builds were working fine.

Suddenly it stopped working. There is nothing in logs

I tried starting it, with the following command :-

sudo /etc/init.d/jenkins start

but still it’s not running

root@localhost:$# service jenkins restart
 * Restarting Jenkins Continuous Integration Server jenkins [ OK ] 

root@localhost:$# service jenkins status
Jenkins Continuous Integration Server is not running

After some googling I was able to get to run jenkins with the foll command :-

-Djava.awt.headless=true -jar /usr/share/jenkins/jenkins.war –webroot=/var/cache/jenkins/war –httpPort=8080

Still not able to get why the conventional Jenkins start command is not working. I’m working on ubuntu 14.04 version. So maybe there is some issue with Jenkins-Linux combination. To get it working is enough for the time being!