Overnight, my host updated my PHP from 5.3.18 to 5.3.20. At the same time, they removed some modules/extensions that were believed to be causing intermittent crashes related to segmentation faults. The suspect extension was ioncube, but some pdo and sqlite modules were also removed as they had been automatically added at a previous update about the time when the intermittent crashes began.
Since the update last night, I've had various issues such as members being blocked due to login quota failures. I disabled all plugins to make sure it wasn't one of my security plugins gone haywire. I had support truncate the strikes table and was able to login and disable the strikes feature temporarily.
I noted that in the active user list, every visitor is reported as having the server's own IP address. This is clearly the cause of the login issues, but I don't know what is causing the IP issue in the first place.
I'm on a dedicated server at FutureHosting that had been optimized by a consultant (a very reputable one). In searching similar reports, some others with the same issue identified some type of proxy like cloudflare as the cause, but I do not believe my server has anything like that unless it was automatically added in the overnight PHP update.
I am running APC and cpanel varnish cache, though they had not been an issue prior to last night's maintenance.
Any idea what may be causing this, particularly something that might have happened in an apache/php update?
Since the update last night, I've had various issues such as members being blocked due to login quota failures. I disabled all plugins to make sure it wasn't one of my security plugins gone haywire. I had support truncate the strikes table and was able to login and disable the strikes feature temporarily.
I noted that in the active user list, every visitor is reported as having the server's own IP address. This is clearly the cause of the login issues, but I don't know what is causing the IP issue in the first place.
I'm on a dedicated server at FutureHosting that had been optimized by a consultant (a very reputable one). In searching similar reports, some others with the same issue identified some type of proxy like cloudflare as the cause, but I do not believe my server has anything like that unless it was automatically added in the overnight PHP update.
I am running APC and cpanel varnish cache, though they had not been an issue prior to last night's maintenance.
Any idea what may be causing this, particularly something that might have happened in an apache/php update?
Comment