I see what you mean. And also, if you're making the change in the web.config, that should trigger an application restart anyway.

Something is preventing the pageSecurity object from being set. You could do a global text search for "pageSecurity" and find the code that sets it. That may be where the failure is happening.

Normally I would expect this to be caused by a missing security-related table, like Roles or something. But you said it was a full copy. You might want to double check.

There you go, that should do it!

@scottalanmiller said:

That Facebook is using HHVM in production makes me a little more comfortable with it.

True, but I would guess that Facebook's application runs mostly on code that Facebook is responsible for maintaining. On the other hand, most WordPress websites have at least a couple of third party plugins and/or themes and those usually assume you're running on PHP 5 so there's a higher chance of stuff like this happening.

Probably wouldn't be an issue very often but I would be more paranoid about updates breaking things if I were running HHVM in production.

Me too, that sounds good now that you say it.

I am staying with my dad too, here in the land of pizza, subs and fish frys.

Trying anything I can think of, this part looks promising to maybe be the issue...

WSREP_SST: [INFO] Removing the sst_in_progress file (20150615 23:31:45.369) 150615 23:31:45 [ERROR] WSREP: Process completed with error: wsrep_sst_xtrabackup --role 'joiner' --address 'x.x.x.x' --auth 'no:no' --datadir '/db/mysql/' --defaults-file '/etc/mysql/my.cnf' --parent '16465': 32 (Broken pipe) 150615 23:31:45 [ERROR] WSREP: Failed to read uuid:seqno from joiner script. 150615 23:31:45 [ERROR] WSREP: SST failed: 32 (Broken pipe) 150615 23:31:45 [ERROR] Aborting

MySQL seems to be completely hosed. I'm going to have to rebuild the server, as restoring from backups even a week ago, even at the VM level, is not working!

@JaredBusch said:

@tonyshowoff said:

Using anything but UTF-8 is a crime!

True, and even old WP installs used it without the actually making sure the DB was configured that way. So odd.

That kind of ineptitude seems par for the course for WordPress. UTF-8 issues can't be blamed just on them though, consider the absolutely bizarre issue of utf8 vs utf8mb4 for MySQL, why the hell wasn't it implemented completely the first time?

@scottalanmiller said:

Often with those choices you get PostgreSQL too, which would be my preference nine times out of ten.

You can go straight to hell. That's right, an old rivalry that I can no longer justify.

@scottalanmiller said:

No amount of IT experience prepares you to be a developer. It's a different job, a different range of skills.

True that. You'll also run into many developers who can't do simple things like install software or configure IP settings on their computer, etc. Two very different fields.

@Ambarishrh said:

@JaredBusch Nice! was not aware of that, so CentOs already switched to MariaDB? 🙂

Yes. RHEL = CentOS.

Most of the Linux distros have switched up MySQL or MariaDB. That is where Oracle is really losing ground. People think that they are using MySQL but under the hood it is MariaDB. Slowly the MySQL marketshare is eroding.

Wow if i look back the day when i first met the CentOS. Thanks to @scottalanmiller and @Minion-Queen because of you guys I learned about them and yeah I will keep going in learning 🙂

I tried adding open-files-limit=4096 in /etc/my.cnf and restart mysql. Then executed automysqlbackup again, this time got something else.

Errors reported during AutoMySQLBackup execution.. Backup failed
Error log below..
-- Warning: Skipping the data of table mysql.event. Specify the --events option explicitly.
mysqldump: Got error: 1142: SELECT,LOCK TABL command denied to user 'root'@'localhost' for table 'cond_instances' when using LOCK TABLES

Checking for fixes