@Jorn Franke In the link you shared, both replace and delete methods returns an exception if the field is not already there. I am wondering how to handle both cases mentioned below.
- When a fresh collection is created, fieldnames and fieldtypes would not already exist. Hence they have to be created. - When the index process is run again and again, fieldnames might already be defined in the schema. In that case creating part has to be ignored. @Shawn It is likely to be the issue. Thanks for sharing. Regards, Arunan *Sugunakumar Arunan* Undergraduate - CSE | UOM Email : aruna <aru...@wso2.com>ns...@cse.mrt.ac.lk Mobile : 0094 766016272 <076%20601%206272> LinkedIn : https://www.linkedin.com/in/arunans23/ On Sun, 20 Jan 2019 at 14:02, Shawn Heisey <apa...@elyograg.org> wrote: > On 1/19/2019 8:47 PM, Arunan Sugunakumar wrote: > > I used SolrJ 7.2.1 and Apache Solr 7.2.1 in one of my projects. Prior to > > data indexing, I create the fields and field types using SchemaAPI in > > SolrJ. If a same field name is repeated, solr did not return an > exception, > > but simply ignored it (or over-rid it). But I tried Solr 7.6.0 with my > > project and it returned exceptions. I went over the Solr logs and found > the > > below error. It seems that I cannot create a field again and again. I > would > > like to know whether I should do something different to overcome this > > problem. > > I'm guessing the changed behavior is a result of this issue, first > released in 7.3.0: > > https://issues.apache.org/jira/browse/SOLR-11631 > > The API call to add a field very likely failed when the field already > existed in 7.2.1, but Solr was returning a zero status, which would tell > the client that it succeeded. Your program would never know there was a > problem. > > Thanks, > Shawn > >
