apachesolr_search failing to index on cron.php or batch index

subscribing

We've seen this happening and (I think) one of the common causes is some "bad" node object causing problems during indexing.

One way to trace this is to reduce the number of nodes indexed per cron call to the minimum and run cron to see when it fails. Then you might have the "bad" node on that batch. The batch is determined by apachesolr_get_nodes_to_index(); you might add some debugging into that function to show what nodes are being sent to be indexed.

Any fatal error will stop the indexing. My guess is that you're running out of memory. Try lowering the number of nodes you're indexing in a cron run (this number also affects the size of the batch operations). You might have to do it in several steps: check the remaining count, try running cron, if it didn't change, lower the number, run cron, etc.

I have similar issue which i came across yesterday whilst trying to update the index.
I have tried both version 1 and 2 but alas cron fails as soon as I enable the apache_search module (without it the cron runs fine albeit no content is indexed.

I tried running the cron code from cli and, it seems to return the array fine but as soon as it is passed to the 'apachesolr_index_nodes' a 'Segmentation fault' occurs :/

Below is the code I used as well as the output

#!/usr/bin/php
<?php
chdir(dirname($_SERVER['argv'][0]));
include_once('./includes/bootstrap.inc');
drupal_bootstrap(DRUPAL_BOOTSTRAP_FULL);
ini_set('memory_limit', '1024M');
ini_set('max_execution_time','180');

$cron_limit =5;
$rows = apachesolr_get_nodes_to_index('apachesolr_search', $cron_limit);
var_dump($rows);
apachesolr_index_nodes($rows, 'apachesolr_search');
?>

The output I get is as follows:

hop@dev:~/public_html/upgrade/htdocs$ ./solrindex.php.sh 
array(5) {
  [0]=>
  object(stdClass)#322 (2) {
    ["nid"]=>
    string(4) "4270"
    ["changed"]=>
    string(10) "1244040112"
  }
  [1]=>
  object(stdClass)#324 (2) {
    ["nid"]=>
    string(4) "4273"
    ["changed"]=>
    string(10) "1244040549"
  }
  [2]=>
  object(stdClass)#325 (2) {
    ["nid"]=>
    string(4) "4268"
    ["changed"]=>
    string(10) "1244040655"
  }
  [3]=>
  object(stdClass)#326 (2) {
    ["nid"]=>
    string(4) "4275"
    ["changed"]=>
    string(10) "1244041409"
  }
  [4]=>
  object(stdClass)#327 (2) {
    ["nid"]=>
    string(4) "4277"
    ["changed"]=>
    string(10) "1244041586"
  }
}
Segmentation fault

My setup is as follows:

Web server:          Apache/2.2.3 (Debian) PHP/5.2.0-8+etch16
MySQL database:	   5.0.32
PHP:	                   5.2.0-8+etch16
PHP memory limit: 512M

See http://drupal.org/node/719356 for the patch that fixed the apachesolr + PHP 5.2.0-8+etch16 problem for me.
"flk" confirmed that that patch worked for him!

Leaving issue open as it is common, but downgrading to normal.

I experienced this issue on D6.16, PHP 5.2.10-2ubuntu6.4 with Suhosin-Patch 0.9.7 on Ubuntu 9.10 - i tried all the above and the link to the patch and nothing worked. It definitely wasnt However, i did find an error relating to a pcre library and upon further investigation this was found to be an issue with the ubuntu compiled php.

So, i purged the ubuntu php, downloaded php-5.2.13, compiled, restarted apache2 (using the stock apache2) and its working fine now.

Hope this helps someone.