Found issues (height of the table above the grid is increased with each window resize), when the mode="minimal" attribute on a scroll container is used.

Removed an empty line.

Updated test plan:

Fixed issue, where database cache without expiration had 0 instead of -1 lifetime stored.

In D443#8991, @erik wrote:

Partially tested for Memcached. Failed test with $application->addCache('key_with_exp', 'v2', 3600);

Searched for new key with command key_memcache_dump.sh key_with

Script scanned 50 slab numbers and stopped. Found only this (lines with empty slabs omitted):

ITEM site_serial:1766886194:1:key_with_exp_serials [12 b; 1679428231 s]
ITEM site_serial:1766886194:1:key_with_def_exp [2 b; 1681584439 s]
Connection closed by foreign host.
Looking for 'key_with' key in 3 slab...
ITEM site_serial:1766886194:1:key_with_def_exp_serials [16 b; 1681584439 s]

Failed test "Testing internal caches expiration for DB cache" in the part "all other keys have an expiration of 25 days (with a few seconds margin) relative to the time they're set".

Partially tested for Memcached. Failed test with $application->addCache('key_with_exp', 'v2', 3600);

In D443#8977, @erik wrote:

Testing internal caches expiration for Memcache
ITEM test [1 b; 0 s]
Failed condition "these keys have infinite expiration (the 1738250904 number might differ on your install)" - given "0 s" is not like some great integer like "1738250904"

some other keys, containing "master", found only with request

(sleep 1; echo "stats cachedump 3 0"; sleep 1; echo "quit";) | telnet localhost 11211 | grep 'master'
response:
ITEM site_serial:1766886194:1:master:cms_menu_serials [15 b; 0 s]
ITEM site_serial:1766886194:1:master:template_mapping_serials [23 b; 0 s]
ITEM site_serial:1766886194:1:master:last_cache_rebuild_serials [25 b; 0 s]
ITEM site_serial:1766886194:1:master:config_files_serials [19 b; 0 s]

keys have "_serials" at the end, so do not match keys from test plan
no other keys with "master" found with other requests

Replicated MemcacheCacheHandler class updates to the MemcachedCacheHandler class.

Testing internal caches expiration for Memcache
ITEM test [1 b; 0 s]
Failed condition "these keys have infinite expiration (the 1738250904 number might differ on your install)" - given "0 s" is not like some great integer like "1738250904"

In D443#8969, @erik wrote:

Testing internal caches expiration for Memcache
ITEM site_serial:1766886194:1:master:last_cache_rebuild [68 b; 0 s]
Failed condition "these keys have infinite expiration (the 1738250904 number might differ on your install)" - given "0 s" is not like some great integer like "1738250904"

Done requested changes - returned fallback to FakeCacheHandler zero time parameterusage and comments.

Testing internal caches expiration for Memcache
ITEM site_serial:1766886194:1:master:last_cache_rebuild [68 b; 0 s]
Failed condition "these keys have infinite expiration (the 1738250904 number might differ on your install)" - given "0 s" is not like some great integer like "1738250904"

In D443#8964, @erik wrote:

Detected that after memcache restart, test setCache with $application->setCache('key_with_def_exp', 'v1');, is passed only when twice open index.php. First index.php opening does not set value.

Detected that after memcache restart, test setCache with $application->setCache('key_with_def_exp', 'v1');, is passed only when twice open index.php. First index.php opening does not set value.

Fixed ArgumentCountError exception on the System Configuration step during clean install on PHP 7.4.

Set infinite expiration for site serial key.

Renoved fallback logic, updated test plan.

1. write a test plan for:
   • memory cache testing
   • database cache testing
   • template cache testing (via cache_timeout tag parameter, m_Cache tag and similar stuff)
2. store site serial with infinite expiration, because otherwise, it's expiration after 1 month will invalidate the whole cache

1. attempts to extract code from the trace, that doesn't contain a file reference
2. code of the source file/line isn't shown (because it's outside of a trace)

Adding phrases.

Fixed outdated phrase translation.

Changed terminology

Need to:

• finalize the plan for dead lock testing
• may add test plan for reconnect testing

Tried test part 4. No messages in system log given. Maybe, because of START TRANSACTION SQL command. Transaction not committed, so no records created in DB.

Bugfix: logging retry attempts reset retry counter resulting in indefinite retry attempts.

In D439#8859, @erik wrote:

@alex trouble in part3 1 G - cant't get "detailed lock retry info is shown in the system log".

On processing line with "Step 4." comment, i have lock much more then 50 seconds. Practically, it never ends. In case, when lock continue, in other debug window go to step 5 - lock ends, but no system log record about this lock created.

In D439#8859, @erik wrote:

@alex trouble in part3 1 G - cant't get "detailed lock retry info is shown in the system log".

@alex trouble in part3 1 G - cant't get "detailed lock retry info is shown in the system log".

Send e-mail upon retry (for real).

Send e-mail upon retry.

Added an explanatory comments.