Oh, I didn't realize that fetch plan is only for remote: and http: access. I don't see this in the wiki: https://github.com/orientechnologies/orientdb/wiki/Fetching-Strategies
Wouldn't this be important for consistency and speed? For consistency, fetching based on a plan in one shot guarantees that everything is consistent as of that point in time. Otherwise, if the graph is walked over say 30 seconds in an actively updated graph, lots of Vs and Es are going to be inconsistent with the first select or am I missing something? For Speed, I would think that getting everything in shot will improve performance b/c 1 read could be done instead of 100s. Even though it's plocal access, there's still a lot of overhead (I don't have proof of this but am assuming there is a substantial amount with larger result sets). On Monday, March 31, 2014 11:49:02 AM UTC-4, Andrey Lomakin wrote: > > HI, > Fetch plan is used in remote and http protocols. > Why do you need it in embedded database ? > > > > On Thu, Mar 27, 2014 at 8:53 PM, cp2 <[email protected] > <javascript:>>wrote: > >> using 1.7rc2 SNAPSHOT from 10 March, 2014 >> >> Either I am doing it wrong or FetchPlans seem to be ignored running >> embedded and plocal. >> If I disable all caching, I get only the local record, >> I use default caching, I get back the entire connected object tree >> >> Tested using the schema and data from the Fetch Plan wiki page >> https://github.com/orientechnologies/orientdb/wiki/Fetching-Strategies >> >> with caching disabled, I always get this for fetch plans: *:-2 *:0 *:1 *:-1 >> and no fetch plan >> *Invoice#11:0{out_customers:#12:0,out_addresses:#13:0,out_orders:[size=3]} >> v7* >> >> with default caching I get for all records regardless of fetch plan: >> >> * >> Invoice#11:0{out_customers:#12:0,out_addresses:#13:0,out_orders:[LineItem#16:0{count:5,in_orders:#11:0,out_order_item:#17:0} >> >> v4, LineItem#16:1{count:3,in_orders:#11:0,out_order_item:#17:1} v4, >> LineItem#16:2{count:4,in_orders:#11:0,out_order_item:#17:2} v4]} v7* >> >> Also, using the debugger, with caching enabled, I can browse the entire >> object graph from the Invoice document -- which I would expect only for >> fetch plan *:-1. >> When I have caching disabled, all I can see are the RIDs of linked >> documents, which I would only expect for fetch plan *:0 or *:-2 >> >> this is my code >> >> 'select from Invoice' >> >> >> >> >> *static List<ODocument> query(OrientGraph pGraph, final String pSql, >> final String pFetchPlan) { final OQueryAbstract zQuery = new >> OSQLSynchQuery(pSql).setFetchPlan(pFetchPlan); return >> pGraph.getRawGraph().query(zQuery);}* >> >> This leaves me confuse about what the fetch plan does. I don't want my >> queries returning the whole graph. What am I doing wrong? >> >> -- >> >> --- >> You received this message because you are subscribed to the Google Groups >> "OrientDB" 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. >> > > > > -- > Best regards, > Andrey Lomakin. > > Orient Technologies > the Company behind OrientDB > > -- --- You received this message because you are subscribed to the Google Groups "OrientDB" 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.
