First I want to thank you for the quick reply. I think instead of wasting more of your time I will have to sit down and learn Neo4j's query language and understand the types of queries as well as the data stored in the database. I am a research student placed in a project where all those who have built the database have left and I am stuck with understanding what's in the database and how to query that information. Considering your point about how I should use graph queries instead of pulling the whole database from disk I believe a good start would be to learn the graph that I am dealing with then understanding the queries that obtain the information I need.
Thank you for your time and have a nice day. JP On Sunday, May 18, 2014 8:43:25 AM UTC-4, Michael Hunger wrote: > > You shouldn't remove information from your query, if need be anonymize it > or adapt it. Otherwise you can't get real help. > > What is the actual use-case you want to solve? What do you actually want > to do? > > The query you had will do a full graph scan, not something you want to do. > > Make sure to use graph queries (i.e. ones that actually traverse > relationships). > > The query you have pulls all data in your database from disk and pushes it > through a Cypher query (i.e. 600M elements), certainly not a real use-case. > > If you split your union into two queries, how long is the runtime of each > of the queries? > > for the second one I'd probably go with start r=rels(*) WHERE ... > > > > On Sun, May 18, 2014 at 6:12 AM, gfourhsean Ionessa < > [email protected] <javascript:>> wrote: > >> Hello Neo4j Guru's, >> >> I know this question has been asked already in previous posts (which I >> have read) but I feel like something else is causing my issue maybe. >> >> To lay out as much as I can, I have been assigned to manage a database >> based off of Neo4j version 2.0.0 with a database size of ~250 Gigabytes. >> The problem my team has been encountering is ridiculously slow database >> querying (from 30 minutes to an hour before receiving results). The queries >> are executed using the web front end on the webadmin page. Below are the >> details for the graph nodes and relationship details to get an ideal of the >> size of the database I am dealing with. I am wondering if the sheer size of >> the database is the cause for such a slow query performance or is there >> something else that may be going on which I must look into. >> >> Test Query (With actual label information removed for privacy reasons): >> >> MATCH (n) WHERE has(n.`person`) RETURN DISTINCT "node" as element, >> n.`person` AS `person` LIMIT 25 UNION ALL MATCH ()-[r]-() WHERE >> has(r.`person`) RETURN DISTINCT "relationship" AS element, r.`person` AS >> `person` LIMIT 25 >> >> *The server Configuration:* >> >> CPU's: 24x E5-2630 2.30 GHz >> Memory: 125 Gigabytes of DDR3 >> Hard disk: 5 Terrabytes >> >> *JVM Settings manually set (in neo4j-wrapper.conf):* >> >> # Initial Java Heap Size (in MB) >> wrapper.java.initmemory=16 >> >> # Maximum Java Heap Size (in MB) >> wrapper.java.maxmemory=32768 >> >> *Linux system settings manually set:* >> >> ulimit set to 40000 >> >> under /etc/sysctl.conf: >> vm.dirty_background_ratio = 50 >> vm.dirty_ratio = 80 >> >> *Node and relation information:* >> >> "description": "Estimates of the numbers of different kinds of Neo4j >> primitives", >> "description": "An estimation of the number of nodes used in this >> Neo4j instance", >> "name": "NumberOfNodeIdsInUse", >> "value": 255857300, >> >> "description": "An estimation of the number of relationships used in >> this Neo4j instance", >> "name": "NumberOfRelationshipIdsInUse", >> "value": 375613862, >> >> "description": "An estimation of the number of properties used in >> this Neo4j instance", >> "name": "NumberOfPropertyIdsInUse", >> "value": 1518707868, >> >> >> "description": "The number of relationship types used in this Neo4j >> instance", >> "name": "NumberOfRelationshipTypeIdsInUse", >> "value": 8, >> >> >> I want to apologize in advance for my lack of knowledge of Neo4j but I am >> willing to learn as much as I can to fix this issue. >> >> Sincerely, >> JP >> >> -- >> 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] <javascript:>. >> 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.
