Details
-
Bug
-
Status: New
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
-
Rank:0|i004an:
-
SQL Server
-
All Web Browsers
Description
I'm not sure if this is a bug or something we have configured incorrectly.
This started by referring to a genomics solution in the i2b2 community: https://community.i2b2.org/wiki/display/IGD/Loading+Genomic+VCF+Files+into+i2b2.
When we use the GENOTYPE datatypes (e.g. GENOTYPE_RSID_SNP), i2b2 displays the nicer search windows. However, after specifying values and clicking “Run Query” in the query tool, it runs 12 seconds, says it timed out, and shows “ERROR” in the Previous Queries panel. We tried running Wildfly in the console and were able to see a vague error message saying “Job Exception: NULL”. It looks like whatever the error is, there isn’t enough info to return an actual message. That error message appears immediately in the console after clicking “Run Query”. We never see the SQL query in the console. So, I’m guessing it never makes it to the db. Here are some other details:
- We originally were using Postgresql and figured that was the issue. So we switched to SQL Server and are still having a problem.
- We are using the SQL Scripts from the above named solution, with a couple small fixes.
- We tried changing C_METADATAXML to use generic LARGESTRING as the datatype, instead of the following datatypes: GENOTYPE_GENE_INDEL, GENOTYPE_GENE_SNP, GENOTYPE_RSID_INDEL, GENOTYPE_RSID_SNP. Using LARGESTRING results in the searches functioning correctly; returning accurate patient counts.
- We also have tried your example .NET etl application for vcf files. That works fine loading the sample vcf info into observation_fact. We also have the PHP script working for auto-populating the search box.
- We’re using i2b2 version 1.7.10, with SQL Server 13.0.5233.0
Based on the fact that our instance of i2b2 works well using LARGSTRING to search via full-text index, I’d say we have the ontology, pm, and crc loaded/configured correctly. That was the only difference in the loading scripts.
This started by referring to a genomics solution in the i2b2 community: https://community.i2b2.org/wiki/display/IGD/Loading+Genomic+VCF+Files+into+i2b2.
When we use the GENOTYPE datatypes (e.g. GENOTYPE_RSID_SNP), i2b2 displays the nicer search windows. However, after specifying values and clicking “Run Query” in the query tool, it runs 12 seconds, says it timed out, and shows “ERROR” in the Previous Queries panel. We tried running Wildfly in the console and were able to see a vague error message saying “Job Exception: NULL”. It looks like whatever the error is, there isn’t enough info to return an actual message. That error message appears immediately in the console after clicking “Run Query”. We never see the SQL query in the console. So, I’m guessing it never makes it to the db. Here are some other details:
- We originally were using Postgresql and figured that was the issue. So we switched to SQL Server and are still having a problem.
- We are using the SQL Scripts from the above named solution, with a couple small fixes.
- We tried changing C_METADATAXML to use generic LARGESTRING as the datatype, instead of the following datatypes: GENOTYPE_GENE_INDEL, GENOTYPE_GENE_SNP, GENOTYPE_RSID_INDEL, GENOTYPE_RSID_SNP. Using LARGESTRING results in the searches functioning correctly; returning accurate patient counts.
- We also have tried your example .NET etl application for vcf files. That works fine loading the sample vcf info into observation_fact. We also have the PHP script working for auto-populating the search box.
- We’re using i2b2 version 1.7.10, with SQL Server 13.0.5233.0
Based on the fact that our instance of i2b2 works well using LARGSTRING to search via full-text index, I’d say we have the ontology, pm, and crc loaded/configured correctly. That was the only difference in the loading scripts.