How to Increase THRESHOLD

Hi @faarukh, thank-you for providing those logs.

I’ve taken a look and the error_log file, which contains the Commands out of sync error from both Wordfence and for a WooCommerce table. There are also “deadlock” errors logged on wp_options in the php_errorlog file, along with MySQL server has gone away reading the result set’s header. WordPress also produces a Server shutdown in progress error when returning some queries.

These suggest that the database server is actually restarting or possibly crashing and starting again automatically. Running out of memory can be a cause for this. If you run the server yourself, you’ll need to check out the MySQL logs and syslog to confirm, otherwise, please get in touch with your host who should be able to check it out for you. If the database server is crashing or running out of memory, they should be able to diagnose and resolve the problem for you.

Thanks again,

Peter.

Hi

Ok, I am sending this reply to my hosting provider let’s see what they can do, and then I will get back to you.

Regards
Farrukh

Thanks @faarukh, let us know how you get on!

Peter.

Hi Peter

Here is the reply from my hosting:

“I went through the Wordfence troubleshooting article but that didn’t seem to help the issue.

Your account resources look ok too at this time when the scan was running but you can also try to disable all the plugins to make sure there is no conflict then try again.

I even turned off all the scanning options so you would think it would just start then complete but that didn’t seem to work either which tells me there may be some conflict.

I just checked all the PHP settings and limits again and they seem great for this to work properly so you may have to get back to their support and see if they can narrow down why it’s failing.

It could be a corrupted file or database entry that may be stalling the process – that is my best guess.”

Please note that I also try by disabling the plugins but the scan failed.

Regards
Farrukh

Waiting for your response.

Hi @faarukh,

The host’s support said that your account resources seem OK, but the issue is that the MySQL server has gone away and related messages mean the whole server may be running low on resources, and it’s likely killing mySQL because that’s the biggest user of memory.

The host is right that there could be database corruption causing bad values to come back from queries, but we feel that would be unlikely to occur on three separate sites at once.

Please can you run a scan, and when it fails, note the date/time. Then find if the timestamps on some of the mySQL error messages match up to the time that it failed.

If we can see some of those actions happening at the same time, then asking the host to see why the “gone away” message happened may help. They can do this by looking at mySQL’s error log or the syslog, and they should find something close to the same timestamp you have identified.

Thanks,

Peter.

Thanks for the details.

I will run the test again and then ask my host to check the details then I will back to you.

Regards
Farrukh

Hi Peter,

Here is the reply from the host:

“I was able to run some tests and troubleshooting for an hour or so and found that there is some corruption or conflicting metadata in the database that is preventing the Wordfence scan from completing.

I made a duplicate of that WordPress installation and connected it to a new database then ran the Wordfence scan again and it completed successfully within about 5 minutes.

It would be best to report this to Wordfence.”

Regards
Farrukh

Waiting for your reply.

Hi @faarukh, I was on annual leave so I apologize for taking a few days to reply.

It sounds like positive news from the host, although I’m not 100% sure of the origin of that metadata in the first instance, but glad the scans can complete on a fresh installation so your server configuration or Wordfence setup doesn’t seem to be faulty.

It seems that the data causing the issue is within your WordPress database, but the cause or specific record is not clear. Did they give you more specific information on data that could be removed or repaired in your existing site’s database? If not, we could potentially request a backup copy of your database to see if we can replicate the error and identify the cause of the problem so that you can remove it. It may also help us identify data to ignore or treat differently when a scan is running.

Thanks again,

Peter.

Hi Peter,

Thanks for the reply.

I will get back to you for further details.

Regards
Farrukh

Hi Peter,

Please tell me how can I send you a copy of the database?

Regards
Farrukh

Hi @faarukh,

You could send a compressed file, or link to online storage such as a private Dropbox link to wftest @ wordfence . com with your forum username in the subject line so I can easily find it.

Let us know when you have sent it and I’ll take a look!

Peter.

Hi Peter,

I have emailed the database of my website.

Subject: @faarukh – Copy of Database

Regards
Farrukh

Hi @faarukh, thanks for your patience, I sent the copy of your database over to our development team.

They’ve taken a look, and the wfconfig table where the currentCronKey value is stored during scans seems fine. It’s empty, which is normal when a scan isn’t in the middle of forking. It’s not possible to tell if it can’t be written properly during a scan just by looking at the export. Looking through the other values, they look normal too.

If it’s a broken index, we would have expected the delete and deactivation explained earlier in the ticket to fix it, but there have been MySQL bugs in the past where a table gets broken but the issue survives re-creating of the table. It may be that kind of issue, but that involves exporting the whole database and re-importing it with a new database name. It’s quite a drastic measure though, especially if we can’t tell in advance whether that would definitely fix the issue.

Thanks,

Peter.