Hi Alan So IIRC, there are outstanding faults with the uniqueness modes that underpin the *add_to_index_or_fail* function. I forget the exact nature of these but I put a warning in the py2neo docs about using them: http://py2neo.org/2.0/legacy.html#py2neo.legacy.LegacyWriteBatch.add_to_index_or_fail
If memory serves (which it often doesn't) the issue was related to this: https://github.com/neo4j/neo4j/issues/906 Going forward, I'd advise moving to labels and schema indexes rather than using legacy indexes. Additionally, Cypher transactions are generally superior to the old HTTP batches which have some functional limitations. Both of these features (schema indexes and Cypher transactions) are now mature and still in active development so will often provide better results. Cheers Nigel On 19 August 2015 at 13:30, Alan Robertson <[email protected]> wrote: > I turned on Nigel's debugging. This is the error this time (same but > different node number mentioned) > BatchError: Batch job 2 failed with NotFoundException > Node 35832 not found > Here's what Nigel's debugging says: > > > Sending batch request with 3 jobs > > {0} POST node > {"domain":"global","nodetype":"Person","firstname":"Annika","lastname":"Hansen","time_create_iso8601":"2015-08-18 > 20:28:46","time_create_ms":1439929726819,"dateofbirth":"unknown"} > > {1} POST {0}/relationships {"to":"node/35910","type":"IS_A"} > > {2} POST index/node/Person?uniqueness=create_or_fail > {"value":"Annika","key":"Hansen","uri":"{0}"} > < Received batch response for 3 jobs > < {0} 201 <Node graph=u'http://localhost:7474/db/data/' ref=u'node/35917' > labels=set([]) properties={u'domain': u'global', u'node > type': u'Person', u'firstname': u'Annika', u'lastname': u'Hansen', > u'time_create_iso8601': u'2015-08-18 20:28:46', u'time_create_ > ms': 1439929726819, u'dateofbirth': u'unknown'}> > < {1} 201 <Relationship graph=u'http://localhost:7474/db/data/' > ref=u'relationship/60444' start=u'node/35917' end=u'node/35910' t > ype=u'IS_A' properties={}> > It points to the problem as being with the index/node/Person update. > > > On 08/17/2015 11:50 AM, Alan Robertson wrote: > > Hi, > > > > Twice I've tried to convert to py2neo 2.x. Both times I ran into this > problem. The first time I had to revert the 2.0 changes and put out the > release anyway. This time I'm sticking with it until I get past it > somehow... > > > > The symptom is that I get this result from a transaction: > > > > BatchError: Batch job 2 failed with NotFoundException > > Node 34950 not found > > > > I can assure you that the number 34950 does not appear anywhere in my > code (I put in lots of debug). The transaction consists of three things: > > Create a new node > > Relate the newly created node to an old node (which has node id > 35031). > > Insert the newly-created node into a legacy index. > > > > This all works with the 1.6 series of code. > > > > Here's a little of the debug from the transaction: > > ==== Performing batch.create(0:({dateofbirth:"unknown",domain:"global", > firstname:"Annika", lastname:"Hansen", > nodetype:"Person",time_create_iso8601:"2015-08-17 17:23:50", > time_create_ms:1439832230360})) - for new node > > ==== RELATIONSHIP(0 -[IS_A]-> (n35031 > {domain:"metadata",name:"Person",nodetype:"CMAclass"})) > > ==== add_to_index_or_fail: node 0; index Index(Node, u' > http://localhost:7474/db/data/index/node/Person')("Hansen","Annika") > > Here's the explanation of the debug output: > > The 0 in "batch.create(0:(..." means that this newly-created node has > index 0 in the transaction (as it should). > > The "RELATIONSHIP(0 -[IS_A]-> (n35031 {..." means that we created a > relationship between a node 0 and a Py2neo Node whose id is 35031. > > The "add_to_index_or_fail" line means we called the > legacy.add_to_index_or_fail function on node 0, putting it into the Person > legacy index with the values "Hansen" and "Annika" as the index values. > > > > This is all exactly the same code which worked before. I mainly added > "legacy." to a bunch of calls, changed a few imports and added more debug > when it didn't work. > > > > This looks a bit like a bug in py2neo to me. Suggestions on how to > proceed? > > -- > > > > Alan Robertson / CTO > > [email protected] / +1 303.947.7999 > > > > Assimilation Systems Limited > > http://AssimilationSystems.com > > > > Twitter Linkedin skype > > -- > > You received this message because you are subscribed to the Google > Groups "Neo4j" group. > > To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected]. > > For more options, visit https://groups.google.com/d/optout. > > > -- > > Alan Robertson / CTO > [email protected] / +1 303.947.7999 > > Assimilation Systems Limited > http://AssimilationSystems.com > > Twitter Linkedin skype > > -- > > Alan Robertson / CTO > [email protected] / +1 303.947.7999 > > Assimilation Systems Limited > http://AssimilationSystems.com > > [image: Twitter] <https://twitter.com/ossalanr> [image: Linkedin] > <https://www.linkedin.com/in/alanr> [image: skype] > <https://htmlsig.com/skype?username=alanr_unix.sh> > > -- > You received this message because you are subscribed to the Google Groups > "Neo4j" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "Neo4j" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
