Hi Luca, first I would like to test ZFS performances using an almost real use case (at least more real than the previous test). Then I'm surely going to test ODB performances in a distributed topology, but I have to wait for creation of loading scripts by my developers team, for OUR real use cases.
Anyway we can have a conversation about this on Friday, during "OrientDB meets Rome". cheers Carlo Il giorno martedì 27 maggio 2014 12:23:55 UTC+2, Lvc@ ha scritto: > > Hi Carlo, > I'm very interested about a new study about performance against ZFS by > using multiple clusters. The best would be one cluster per disk to see how > much improvement we have following this approach. > > We know we still have some non necessary locks in the PLocal Storage & > DiskCache layers that will not allow to squeeze multi-core at the maximum, > but it's something Andrey is working now for the 2.0. > > Lvc@ > > > > On 27 May 2014 10:00, Carlo Pradissitto <[email protected]<javascript:> > > wrote: > >> Hi Riccardo, >> I'm still investigating OrientDB performances on ZFS filesystem. >> Anyway, you can send me a private message, and maybe we can help each >> other. >> >> cheers >> Carlo >> >> >> Il giorno martedì 27 maggio 2014 09:10:45 UTC+2, Riccardo Tasso ha >> scritto: >>> >>> Great, I'm really interested in the boost of performance gained with >>> PARALLEL keyword. >>> >>> Cheers, >>> Riccardo >>> >>> >>> 2014-05-27 9:03 GMT+02:00 Carlo Pradissitto <[email protected]>: >>> >>>> Time to document another performance test >>>> stay tuned! >>>> >>>> >>>> Il giorno martedì 27 maggio 2014 02:02:36 UTC+2, Lvc@ ha scritto: >>>> >>>>> Breaking news: OrientDB 1.7 is out! We made OrientDB faster than >>>>> before and with new exciting features like Distributed Sharding, the >>>>> support for Lucene indexes (Full-Text and GEO-Spatial), SSL connections, >>>>> Parallel queries and more. >>>>> >>>>> To download OrientDB 1.7 go to: http://www.orientechnologies.com/ >>>>> download/ >>>>> >>>>> *What’s new?* >>>>> Core >>>>> >>>>> - New “*minimumclusters*” to auto-create X clusters per class >>>>> - New *cluster strategy* to pick the cluster. Available >>>>> *round-robin*, *default* and *balanced* >>>>> - Added *record locking* via API >>>>> - Removed rw/locks on schema and index manager >>>>> - Cached most used users and roles in RAM (configurable) >>>>> >>>>> Server >>>>> >>>>> - New support for *SSL* connections on binary and HTTP protocols >>>>> >>>>> Studio >>>>> >>>>> >>>>> >>>>> >>>>> * - Bookmarks for queries - Ability to interrupt running commands * >>>>> Console >>>>> >>>>> - New “*script*” command to execute multiple SQL statements + >>>>> transaction control >>>>> - New “*next*” command to move the cursor on the next record in >>>>> the result set >>>>> - New “*pref*” command to move the cursor on the previous record >>>>> in the result set >>>>> - New “*move* ” command to move from current record by executing a >>>>> predicate >>>>> - New “*eval* ” command to evaluate a predicate against current >>>>> record >>>>> - Ability to ignore errors by setting “set ignore = true” >>>>> >>>>> Distributed >>>>> >>>>> - New support for *Sharding* >>>>> - Simplified (JSON) configuration >>>>> >>>>> Graph >>>>> >>>>> - New functions to return the traversed items: *traversedVertex*(), >>>>> *traversedEdge*() and *traversedElement*() >>>>> - Method* addEdge*() can takes ORID to use implicitly last record >>>>> version >>>>> >>>>> SQL >>>>> >>>>> - New *PARALLEL* keyword to run the query on multiple threads >>>>> - New “INSERT INTO … SELECT” to copy records >>>>> - New “INSERT INTO … RETURN ” to return expressions on INSERT >>>>> - New “UPDATE … RETURN ” to return expressions on UPDATE >>>>> - Function distance() accepts measure unit (default = km) >>>>> >>>>> HTTP >>>>> >>>>> - Added support for *context variables* in SCRIPT command >>>>> >>>>> Indexes >>>>> >>>>> - Added *Full-Text Lucene* indexes (as plugin) >>>>> - Added *GEO Spatial Lucene* indexes (as plugin) >>>>> - Added WAL (Journal) support for *Hash-Index* so can be used in >>>>> Transactions >>>>> - Added *NULL* support for composite and non-composite indexes >>>>> - Full-Text can be configured through *METADATA* keyword >>>>> - Full-Text now indexes also sub-words >>>>> - Improved SQL optimizer by avoiding calling evaluate() when not >>>>> needed >>>>> - LINKSET types can be indexed now >>>>> - RID can be used on composite indexes >>>>> >>>>> Full list of all 135 closed >>>>> issues<https://github.com/orientechnologies/orientdb/issues?labels=&milestone=16&page=1&state=closed> >>>>> . >>>>> >>>>> *Enterprise Edition >>>>> <http://www.orientechnologies.com/orientdb-enterprise/>* >>>>> >>>>> OrientDB Community Edition is FREE for any usage (Apache2 license). If >>>>> you want additional features and Professional Support, request a trial of >>>>> Enterprise Edition: Enterprise >>>>> Edition<http://www.orientechnologies.com/orientdb-enterprise/> >>>>> . >>>>> >>>>> >>>>> >>>>> Have fun with Graphs, >>>>> Luca Garulli >>>>> CEO at Orient Technologies >>>>> the Company behind OrientDB >>>>> http://about.me/luca.garulli >>>>> >>>>> - See more at: http://www.orientechnologies.com/orientdb-1-7-is-out/# >>>>> sthash.9ooT15mc.dpuf >>>>> >>>>> -- >>>> >>>> --- >>>> 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. >>>> >>> >>> -- >> >> --- >> 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. >> > > -- --- 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.
