I'm testing pgrouting functions in a tiny server with 2GB of RAM and a dual core processor. I wrote a program that create client process and they connect to the server in a specific hour to perform a query, in fact the query is a pgrouting function call. My trouble is that after 50 process, the server crash and report this error in the postgres log:

terminating connection because of crash of another server process 2012-09-05 15:59:09 VET DETAIL: The postmaster has commanded this server process to roll back the current transaction and exit, because another server process exited abnormally and possibly corrupted shared memory. 2012-09-05 15:59:09 VET HINT: In a moment you should be able to reconnect to the database and repeat your command.

and this error occurs for every process.

The table that contains the graph information has about 300,000 rows. The graph has 250 nodes aprox.

My question: is it ok that this behavior occurs with a server like this and using a table with this amount of rows?

  • Adding the SQL queries you use to the question could be helpful, since maybe someone can optimise them. – lynxlynxlynx Sep 5 '12 at 21:45
  • I tried with a query like this: – José Alejandro Sep 5 '12 at 23:25
  • I tried with a query like this: select shortest_path_astar('select id, source, target, cost2 as cost, x1, y1, x2, y2, reverse_cost2 as reverse_cost from graph_streets', 15478, 56874, true, true); The testing program was written in python using psycopg2. I tried with threads and process. – José Alejandro Sep 5 '12 at 23:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.