alancc's questions

We are using PHP + FPM 8.2 + OPCache. In the past, we use https://www.repairrar.com/opcache.php to check its status. But today it returns an empty page. Although it shows an empty page, I view the page source and find its contents are about 3.3MB, but they are just not show in the browser. I tried Chrome and FireFox, but show empty contents.

I also write a small PHP file

<?php
    echo '<pre>' , var_dump(opcache_get_status(false)) , '</pre>';
?>

When visit https://www.repairrar.com/opcache-get-status.php, it shows the OpCache is enabled and performed properly:

array(8) {
  ["opcache_enabled"]=>
  bool(true)
  ["cache_full"]=>
  bool(false)
  ["restart_pending"]=>
  bool(false)
  ["restart_in_progress"]=>
  bool(false)
  ["memory_usage"]=>
  array(4) {
    ["used_memory"]=>
    int(446911376)
    ["free_memory"]=>
    int(626183176)
    ["wasted_memory"]=>
    int(647272)
    ["current_wasted_percentage"]=>
    float(0.0602819025516510009765625)
  }
  ["interned_strings_usage"]=>
  array(4) {
    ["buffer_size"]=>
    int(134217728)
    ["used_memory"]=>
    int(52098200)
    ["free_memory"]=>
    int(82119528)
    ["number_of_strings"]=>
    int(242556)
  }
  ["opcache_statistics"]=>
  array(13) {
    ["num_cached_scripts"]=>
    int(10741)
    ["num_cached_keys"]=>
    int(15444)
    ["max_cached_keys"]=>
    int(130987)
    ["hits"]=>
    int(167134669)
    ["start_time"]=>
    int(1733489514)
    ["last_restart_time"]=>
    int(0)
    ["oom_restarts"]=>
    int(0)
    ["hash_restarts"]=>
    int(0)
    ["manual_restarts"]=>
    int(0)
    ["misses"]=>
    int(11569)
    ["blacklist_misses"]=>
    int(0)
    ["blacklist_miss_ratio"]=>
    float(0)
    ["opcache_hit_rate"]=>
    float(99.99307851607164820961770601570606231689453125)
  }
  ["jit"]=>
  array(7) {
    ["enabled"]=>
    bool(false)
    ["on"]=>
    bool(false)
    ["kind"]=>
    int(5)
    ["opt_level"]=>
    int(4)
    ["opt_flags"]=>
    int(6)
    ["buffer_size"]=>
    int(0)
    ["buffer_free"]=>
    int(0)
  }
}

And below is the php.ini file(I check the change history and it has not be changed since Opcache.php return visible contents):

; cPanel-generated php ini directives, do not edit
; Manual editing of this file may result in unexpected behavior.
; To make changes to this file, use the cPanel MultiPHP INI Editor (Home >> Software >> MultiPHP INI Editor)
; For more information, read our documentation (https://go.cpanel.net/EA4ModifyINI)

allow_url_fopen = On
allow_url_include = Off
display_errors = Off
enable_dl = Off
file_uploads = On
max_execution_time = 1000
max_input_time = 1000
max_input_vars = 1000
memory_limit = 512M
post_max_size = 64M
session.gc_maxlifetime = 2880
#session.save_path = "/var/cpanel/php/sessions/ea-php72"
upload_max_filesize = 64M
zlib.output_compression = Off
session.save_path = "/var/cpanel/php/sessions/ea-php82"

; The gloabal 10-opcahce.ini has already enable opcache, so it is not necessary to enable it here again.
; Enable opcache here will cause 'Zend OPcache can't be temporary enabled' warning in PHP. So the following
; line is commented out
; opcache.enable=1

I write the following .htaccess rules to redirect all URLs from www.a.com to www.b.com:

RewriteEngine on
RewriteCond %{http_host} ^www.a.com [NC,OR]
RewriteCond %{http_host} ^a.com [NC]
RewriteRule ^(.*)$ https://www.b.com/myfolder/ [R=301,NC,L]

Now there is a file under a subdir of a.com, as below:

www.a.com/.well-known/acme-challenge/#######

It is used for AutoSSL renewal. With the above redirection, the AutoSSL will not be able to find the file any more, so fail the renewal.

So, is it possible to disable the redirect under /.well-known/acme-challenge/?

Today I suddenly find the WordPress Admin dashboard is very very slow. It takes about 60s or more to access the post list, plugin list, etc. Or will just show an time-out error.

I try the following methods:

• Restart Apache service.
• Restart server.
• Change PHP.ini settings. Then restart the server.

But all not working. Finally I do the following steps:

1. Rename .htaccess to .htaccess.backup.
2. Then the dashboard is fast. But WP Rocket complains that it cannot access .htaccess. Note: Based on my test, after renaming .htaccess, I must access a page, then go to step 3 will solve the issue. If Skip step 2, then the speed will not change. It seems that by accessing the page, the .htaccess will be loaded to Apache.
3. Rename .htaccess.backup to .htaccess.

Then the dashboard is still fast, though the contents of .htaccess has never been changed.

Why? I cannot understand why a rename of .htaccess temporarliy can solve the issue permanently, even if the .htaccess contents are actually not changed at all.

Update:

A follow-up update:

1. The problem occurs since I try to install vBulletin in a subfolder of our website https://www.example.com/forums/
2. When the problem occurs, all WordPress instances under the same domain, i.e., https://www.example.com, https://www.example.com/blogs/ and https://www.example.com/license/ are all very slow. However, WP instances under other domains(but in the same server) are NOT affected.
3. After I backup and deleting the /forums/ subdirectory, the problem never occurs again.

Update2:

After one day of normal status, now the problem occurs again. And the only solution is to rename .htaccess to .htaccess.backup, access a page under the domain example.com, and then rename .htaccess back.

Update 3

I finally solve the problem. The detailed process is put at https://www.webmasterworld.com/apache/5088830.htm#msg5089206

Whether I need to escape the slash '/' in RewriteCond?

Currently I write the following rule in .htaccess:

RewriteCond %{QUERY_STRING} rp=/knowledgebase/
RewriteRule ^index\.php$ https://www.datanumen.com/knowledgebase/ [QSD,R=301,L,NC]

However, this only works for URL like https://www.datanumen.com/fi/customer/index.php?rp=/knowledgebase/7/How-to-order-the-full-version-of-DataNumen-Access-Repair.html&language=swedish, but not work for URL like https://www.datanumen.com/fi/customer/index.php?rp=%2Fknowledgebase%2F7%2FHow-to-order-the-full-version-of-DataNumen-Access-Repair.html&language=swedish

So, I have to modify the rule, as below:

RewriteCond %{QUERY_STRING} rp=/knowledgebase/ [OR]
RewriteCond %{QUERY_STRING} rp=%2Fknowledgebase%2F
RewriteRule ^index\.php$ https://www.datanumen.com/knowledgebase/ [QSD,R=301,L,NC]

But I check https://serverfault.com/a/968916/280923 and it said "The slash (/) does not need to be escaped". So I am confused.

If I need to take all situations into consideration, i.e., the escaped version and unescaped version of '/', then there should be totally 4 combination, should I add all of them as the RewriteCond:

rp=/knowledgebase/
rp=%2Fknowledgebase%2F
rp=%2Fknowledgebase/
rp=/knowledgebase%2F

I am launching an EC2 instance and it seems the most popular WordPress AMI in the marketplace is the one provided by Bitnami. However, there is a big issue with the software/system update.

I ask Bitnami and they said if I want to update LAMP on the server, then I need to launch another EC2 instance, then migrate the data from the current instance to the new one. That is really a headache. As I need to do all setups again in the new instance.

I notice that Bitnami AMI will update every day. Does that means I need to do the migration every day to make sure my system contains the up-to-date security updates?

In the past, I use a managed dedicated server on InMotion Hosting, CentOS 6.x + LAMP, and the LAMP will get updated automatically without having to migrate a new server every day.

Therefore, I just wonder for EC2 instance, is there a better way to update the software, including the OS, than doing migration every day?

Thanks

Update

Let me describe the detailed steps to migrate an EC2 instance:

1. Launch a new EC2 instance.
2. Perform all the setups done on the old instance, such as install the SSL certificate, set file/directory permissions, set cron jobs, etc.
3. Copy all WordPress files from the old instance to the new instance.
4. Export the database in the old instance, then import it to the new instance.

The migration of the WordPress can be found at https://wordpress.org/support/article/moving-wordpress/. The migration of other setups are application specific.

Update 2

This https://community.bitnami.com/t/is-the-wordpress-certified-by-bitnami-and-automattic-amazon-image-auto-updated/99631/2?u=ccw is the official support from bitnami and what they suggest is

WordPress can be updated using the admin dashboard, but for the infrastructure components, you will need to launch a new instance and migrate the data when needed

I have checked https://www.parkmycloud.com/blog/aws-regions-and-availability-zones/ and https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using-regions-availability-zones.html#concepts-availability-zones. It is easy to choose a region close to my customers. However, how about the available zone?

Which available zone should I choose when launching an EC2 instance to obtain the best performance? For example, I want to launch an EC2 instance in us-east-1, and then add Cloudfront with Origin shield at us-east1, which available zone should I choose? There are 6 of them. Both articles above do not describe how to choose the available zone.

I notice that in https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using-regions-availability-zones.html#concepts-availability-zones, it said us-east-1a in one account may have the same availalize zone with us-east-1b in another account. Does that mean it is better to let Amazon allocate AZ for me instead of choosing one manually?

I launch an EC2 instance with an AMI from the marketplace, which is called LAMP packaged by Bitnami.

After the instance is launched, I find I can only access its DNS name or IP via HTTP, not HTTPS. It seems that the SSL will not be installed by default.

So I search on its document and find this: https://docs.bitnami.com/aws/faq/administration/generate-configure-certificate-letsencrypt/

I follow the instruction to run the following command:

sudo /opt/bitnami/bncert-tool

But after inputting the DNS name ec2-3-95-160-86.compute-1.amazonaws.com, it shows a warning:

Warning: The domain 'ec2-3-95-160-86.compute-1.amazonaws.com' resolves to a 
different IP address than the one detected for this machine, which is 
'3.95.160.86'. Please fix its DNS entries or remove it. For more info see: 
https://docs.bitnami.com/general/faq/configuration/configure-custom-domain/

I doube-check the domain DNS at https://dnschecker.org/ and confirm ec2-3-95-160-86.compute-1.amazonaws.com will resolve to 3.95.160.86, as below:

So why I still get this warning? Also this warning will prevent me from creating the SSL certificate for ec2-3-95-160-86.compute-1.amazonaws.com at all.

I am following the guide in https://aws-quickstart.s3.amazonaws.com/quickstart-bitnami-wordpress/doc/wordpress-high-availability-by-bitnami-on-the-aws-cloud.pdf to create a WordPress High Availability on AWS.

The creation succeeds. And I can visit the WordPress site and the admin panel. However, when I follow the instructions in section "Option 2. Connect to the cluster through SSH", I find I cannot connect to the Bastion host instance. I always get the error "Connection timed out".

Since I am using SecureCRT, not ssh or PuTTY, so I doubt that there may be some misconfiguration in SecureCRT or it does not work for Amazon EC2 instances. So I try to create a test EC2 instance, and then find I can connect to it with SecureCRT successfully, with the key pair same as that for the Bastion host.

So what I have done wrong? Or what log can I check to solve this issue?

I am using CentOS 6.8.

I have an executable called railgun, whose user and group are both set to root.

I then use the root user to start it as a service, as below:

service railgun start

However, it produces a permission denied error:

/tmp/memcached.sock: connect: permission denied

I check the user & group for /tmp/memcached.sock and find both are "memcached". So I just wonder why railgun cannot access /tmp/memcached.sock since it is executed by root? Which permissions does it have when it is launched by root?

In my opinion, root can access anything. So the executable started by root also can access anything. Is that correct?

I write the following rewrite rule:

<IfModule mod_rewrite.c>
RewriteCond %{QUERY_STRING} rp=\/knowledgebase\/.*
RewriteRule ^\/customer\/index.php /knowledgebase/ [R=301,L]
RewriteRule ^\/customer\/knowledgebase\.php$ /knowledgebase/ [R=301,L,QSA]
</IfModule>

To redirect URL such as

https://www.example.com/customer/index.php?rp=/knowledgebase/5/DataNumen-Excel-Repair to https://www.example.com/knowledgebase/

And redirect URL such as

https://www.example.com/customer/knowledgebase.php to https://www.example.com/knowledgebase/

But both do not work. Why?

Update

I try to put MrWhite's codes from /.htaccess to /customer/.htaccess and make some minor changes to adopt the changes, as below:

RewriteCond %{QUERY_STRING} rp=/knowledgebase/
RewriteRule ^index\.php$ https://www.example.com/knowledgebase/ [QSD,R=301,L,NC]

RewriteRule ^knowledgebase\.php$ https://www.example.com/knowledgebase/ [R=301,L,NC]

Now the redirect works. However, it will only work for case like:

https://www.example.com/customer/index.php?rp=/knowledgebase/9/DataNumen-PDF-Repair

but for case like

https://www.example.com/customer/index.php?a=b&c=d&rp=/knowledgebase/9/DataNumen-PDF-Repair

It will not work. Even after I change ^rp= to rp in RewriteCond.

I just get a Linux server from a third-party. Then I create a SSH tunnel via SecureCRT: https://www.vandyke.com/support/tips/socksproxy.html , where the Linux server is used as the Gateway Server.

However, the Linux server is not 100% safe since it is managed by a third-party. Now if I connect to a HTTPS server via the SSH tunnel, will the password sent to the HTTPS server be stolen by the administrators of the Linux server?

I have a website using PHP-based CMS system.

So each page is generated dynamically, like this :

http://www.shuling.net.cn/index.php?_m=mod_product&_a=view&p_id=87

Now I want to redirect this specific page to another page, i.e.:

• https://www.datanumen.com/sql-recovery/

So I write .htaccess file as follows :

redirect 301 /index.php?_m=mod_product&_a=view&p_id=87 https://www.datanumen.com/sql-recovery/

But after I upload the .htaccess to the root directory of the website, this does not work at all.

I retest by writing another rule, like this:

redirect 301 /sql/ https://www.datanumen.com/sql-recovery/

Then this rule will work perfectly.

Why?

Thank you