Hi- I started seeing this error message today (seems to be generated through the delta index)
Sphinx 1.10-beta (r2420)
Copyright (c) 2001-2010, Andrew Aksyonoff
Copyright (c) 2008-2010, Sphinx Technologies Inc (http://sphinxsearch.com)
using config file
'/mnt/local/flying-sphinx/0b6aa9e26e4765270/sphinx.conf'...
indexing index 'product_search_delta'...
ERROR: index 'product_search_delta': sql_connect: FATAL: too many
connections for role "u6csnjjiujeaas"
(DSN=pgsql://u6csnjjiujeaas:***@127.0.0.1:10361/du3tkbnm0qiva
). total 0 docs, 0 bytes
total 0.114 sec, 0 bytes/sec, 0.00 docs/sec
total 0 reads, 0.000 sec, 0.0 kb/call avg, 0.0 msec/call avg
total 0 writes, 0.000 sec, 0.0 kb/call avg, 0.0 msec/call avg
Any ideas what this might be? We have the ceramic plan running, and a manual rake fs:index seems to work fine (no error messages). The delta indexing had also been working fine for a while, until today when we started receiving these messages. Is this related to the number of connections to our Postgres DB on Heroku, or a sphinx-related error? I could not see anything weird in our logs and a rake fs:index seems to be fine.
#1 Posted by Pat Allan Support staff on 26 Jul, 2012 08:11 PM
Hi Matt
My understanding of the error is that it's related to PostgreSQL - there's too many connections to your database. Perhaps it could be related to the number of workers you're running, both web and job processes? Each delta index job will require a connection, and if there's more of them running at any one time, plus each web worker, plus other tools, I guess that could add up…
So, how many workers are you running, and how often are you seeing delta jobs occur (and how often do they fail, too)?
Odd that it only crops up with the delta indexing… I'd expect to see it with the full index task occasionally as well.
Cheers
Pat