Misc Issues
-
Hello, let me report some things I have noticed that might prove helpful. Some of the things below apply to ‘Info’ plugin.
- When WP rocket is active, there is always a notice ‘atec Cache APCu: Another ‘advanced-cache.php’ already exists – please deactivate it first.’. Even if not using your plugin’s page cache.
- In Opcache, When Consistency is Off (as recommended) the details table shows ‘Off’ in red. Perhaps it should be green.
- The ‘try to raise limit’ always appears. When I increase it, it still appears with a larger number every time.
- In my case there are almost 50% misses. How can i improve this?
Here are my settings:
[opcache]
opcache.enable=1
opcache.memory_consumption=192
opcache.interned_strings_buffer=16
opcache.max_accelerated_files=50000
opcache.max_wasted_percentage=15
opcache.validate_timestamps=0
opcache.mmap_base = 0x20000000
opcache.revalidate_freq=60
opcache.enable_cli=0
opcache.save_comments=0
opcache.load_comments=0
opcache.revalidate_path=1
opcache.use_cwd=1
opcache.fast_shutdown=1
opcache.jit_buffer_size=500000000
opcache.jit=1235
[apcu]
apc.enabled=1
apc.shm_segments=1
apc.shm_size=128M
apc.max_file_size=1M
apc.stat=1
apc.ttl=3600
-
Of course I did 🙂
Anyway this is not an important issue. But let me mention some other things.
When I was talking about misses I was referring to apcu, not opcache (and one suggestion: It would be nice if you also showed fragmentation in your info)
Also in your info plugin, opcache is shown as zero, when it is 512M (memory 0, used 0, free 249M).
Here is my updated config.
[opcache]
opcache.enable=1
opcache.memory_consumption=512M
opcache.interned_strings_buffer=64
opcache.max_accelerated_files=50000
opcache.max_wasted_percentage=10
opcache.validate_timestamps=0
opcache.revalidate_freq=60
opcache.enable_cli=0
opcache.save_comments=0
opcache.load_comments=0
opcache.revalidate_path=1
opcache.use_cwd=1
opcache.jit_buffer_size=1M
opcache.jit=1235
[apcu]
apc.enabled=1
apc.shm_segments=1
apc.shm_size=128M
apc.max_file_size=1M
apc.stat=1
apc.ttl=3600I already use 1.8.31, and I have also tried it without unit. Same output.
Any comment on how to adjust config to reduce misses? I notice that after a while there is a reduction, but it is still too high (approx 35%). Sorry if this is not actually your role (to tell us how to use apcu) but you obviously are far more experienced on this than most of us.
Have a merry Christmas and happy new year !
We need to be precise in what we are talking about.
If you changed opache settings and mem is still 0 you most likely have another .ini overwriting settings. Try atec-system-info and check which settings are effective.
Also flush the opcache so scripts get reloaded.As for APCu, if you mean cache misses – that depends on the cache usage. Are you using it as an object cache or for storage? Or both?
A miss only means that data is not stored but requested. So if some script run apcu_fetch(“foo”) but never apcu_store(“foo”, “bar”) … that will always result in high misses.
But the main issue is most likely that APCu is per PHP worker not per machine. If you have multiple workers running, you might hit different cache per request, thus you will get different results and only over time reach a higher hit rate – if ever.
Remember: High APCu miss rates are a natural side-effect of per-worker caches, not a sign of misconfiguration.
That said, your 128M may be too high, depending on your pm.max_children and machine memory.
MC to you too.
You must be logged in to reply to this topic.