Re: Segmentation fault when running queries in sequence

He Matt
Most of time segment fault is because of broken index, reindex the table (or if you can whole db) it might help get rid of segment faults

Regards
Abbas

> On 23 Sep 2023, at 2:53 PM, Matt Gibbins <matt_gibbins(at)fastmail(dot)com(dot)au> wrote:
>
> 
> I am running postgresql through an SSH tunnel to process a collection of files on the server loading them into the database.
>
> Specifically GTFS files
>
> The process is as follows.
>
> find ../Database\ management/SQL/Create\ tables/Landing -type f -name 'Bus *.sql' | sort -n | while read -r f;
> do
> echo "$f";
> ./Query\ to\ server "$f";
> ./Query\ to\ server "../Database management/SQL/Create tables/AllServices/Insert landing - bus.sql";
> done
> The 'Query to server' script initiates an SSL session with the following statement where postgresql_tcanalysis initiates an SSH tunnel to the database on the remote server which I have SSH connection to.
>
> postgresql_tcanalysis < "$1"
> There is no problem with the SSH connection and running individual queries.
>
> The issue occurs when stepping through the queries.
>
> The find statement results in 19 queries. It is necessary to have individual queries as the content of the GTFS files are not consistent with regard to fields.
>
> Attached is the STDOUT messages showing the progress of the the process to segmentation fault.
>
> Running the individual queries where the segfault occurs does not occur for the individual query. The segfault only occurs when doing the bulk process.
>
> Also attached is the section from /var/log/postgresql/postgresql-15-main.log relating to the segfault occurence.
>
> Any suggestions for identifying the cause appreciated.
>
> Regards
> Matt.
>
>
>
>
>
> <messages.txt>
> <postgresql-15-main-segfault.log>