5.11.2 Confirmation method of event log of Explicit Memory Management functionality
This subsection describes the reference method of event log output by the Explicit Memory Management functionality.
- Organization of this subsection
(1) Confirmation method of output trigger
Events that trigger the log output are output in [cause:<CAUSE>] format. You can confirm the event triggered by confirming the item.
The following table describes the meaning of character string output to <CAUSE>.
<CAUSE> |
Meaning |
---|---|
GC |
Copy GC occurrence |
Full GC |
Full GC occurrence |
Reclaim |
Explicit memory block explicit release process |
Reclaiming |
Java heap overflow in Explicit memory block explicit release process |
New |
Object generation in Explicit memory block |
Migrate |
Automatic release processing for the explicit memory block |
Migrating |
Java heap overflow in the automatic release processing for the explicit memory block |
However, <CAUSE> is not output in a part of the log.
(2) Checking methods using log prefix
You can use log prefixes such as ENS, EVO to perform checking by filtering the log.
The meaning of prefixes of the Explicit Memory Management functionality is as follows:
-
"E" of first character indicates that a prefix is for the log of Explicit Memory Management functionality.
-
Second character indicates log output level. N is output for normal level, V is output for verbose level and D is output for debug level.
-
The third character indicates whether the memory size of Explicit heap or Java heap will change. The memory size changes if the third character is S. The memory size does not change when third character is O. When O, the names of events occurred are output to log. Also, A indicates the log of the automatic allocation configuration file for the explicit memory management.
For example, when you want to check changes of each heap size, perform filtering by S and when you want to check the occurred event, perform filtering by O.
(3) Signs used in description of output format of event log
The following table describes the signs used in the description of output format.
Signs |
Usage examples |
Meaning |
---|---|---|
* |
X* |
Repeat the left side items for minimum zero times. Usage example indicates that X is repeated for minimum zero times. |
? |
X? |
Repeat the left side items for minimum one time. Usage example indicates that X is repeated for minimum once. |
{n,m} |
X{1,5} |
Repeat the left side items for minimum n times and maximum m times. Usage example indicates that X is repeated minimum once and maximum five times. |
{ |
{ABC}* |
The range included in { and } is reference unit at left side of *, ?, {n,m}. Usage example indicates that ABC is repeated for minimum zero times. |
} |
||
\ |
None |
This sign indicates the places with linefeed for visibility in the manual. There is no linefeed in the actual output contents. |
| |
X|Y |
This sign indicates either the left side or the right side. Usage example indicates either X or Y. |
. |
None |
Indicates any character. |
... |
X|Y... |
Repeat minimum once from the line that has same indent as this line and that is nearest to forward match of this line, to the line previous to this line. Usage example indicates that either of X or Y is repeated minimum once. |
In the description of output format, numeric values are expressed by combining signs as in Table 5-32 and signs in Table 5-33.
The following table describes the signs used to indicate numeric values.
Sign |
Definition |
Meaning |
---|---|---|
<digit> |
0|1|2|3|4|5|6|7|8|9 |
Indicates 0 to 9. |
<hex> |
<digit>|a|b|c|d|e|f |
Indicates hexadecimals. |
<const> |
<const><digit>|<digit> |
Indicates positive whole number. |
<decimal> |
<const>.<digit> |
Indicates positive integer (Up to 1st Decimal position). |
<ptr64> |
0x<hex><hex><hex><hex><hex><hex><hex><hex><hex><hex><hex><hex><hex><hex><hex><hex> |
Indicates 64 bit pointer value. |
<ptr32> |
0x<hex><hex><hex><hex><hex><hex><hex><hex> |
Indicates 32 bit pointer value. |
<ptr> |
<ptr64>|<ptr32> |
Indicates pointer value. |
<letters> |
.? |
Indicates any character or character string. |
Note The signs indicated in Table 5-32 would be the reference unit at left side or right side of *, ? or |. For example, when <digit>|<hex> is mentioned, indicates numeric value from zero to nine or hexadecimal.