NUMBUFG parameter: Difference between revisions
m (→See also: slight change to link destination) |
|||
| (3 intermediate revisions by one other user not shown) | |||
| Line 29: | Line 29: | ||
</li> | </li> | ||
</ul> | </ul> | ||
===Setting NUMBUFG=== | |||
<p> | <p> | ||
The maximum number of buffers is <code>2**31/3</code>, or <code>715,827,882</code>, which limits buffer pools to 4.2 terabytes or less.</p> | The maximum number of buffers is <code>2**31/3</code>, or <code>715,827,882</code>, which limits buffer pools to 4.2 terabytes or less.</p> | ||
| Line 36: | Line 38: | ||
</p> | </p> | ||
<p> | <p> | ||
If you set <var>NUMBUFG</var> to a lower value | If you set <var>NUMBUFG</var> to a value lower than the minimum, the value is reset to the calculated value and you receive messages like the following: </p> | ||
<p class="code">*** M204.2756: NUMBUFG reset to NLRUQG * ((NSERVS + NSUBTKS) * MAXOBUF + 15) | <p class="code">*** M204.2756: NUMBUFG reset to NLRUQG * ((NSERVS + NSUBTKS) * MAXOBUF + 15) | ||
= 1 * ((2+9) * 4 + 15) = 59 | = 1 * ((2+9) * 4 + 15) = 59 | ||
*** M204.0075: Initialization completed. Buffers = 0+59 | *** M204.0075: Initialization completed. Buffers = 0+59 | ||
</p> | </p> | ||
<blockquote class="note"><b>Note:</b> | |||
<p>If you set <var>NUMBUFG</var> greater than 0: </p> | |||
<ul> | |||
<li>The X'02' bit of the <var>[[XMEMOPT parameter|XMEMOPT]]</var> parameter (the default), must also be set. </li> | |||
<li>The [[File Load utility: FLOD and FILELOAD commands|FLOD and FILELOAD commands]] and [[Fast/Reload]] are not available due to 64-bit addressing conflicts. This particular restriction is removed in V7.7 and later releases. | |||
</ul> | |||
</blockquote> | |||
==See also== | ==See also== | ||
Latest revision as of 14:49, 26 August 2020
Number of above-the-bar buffers
Summary
- Default value
- 0
- Parameter type
- System
- Where set
- Not resettable
- Related products
- All
- Introduced
- Model 204 V7.1
Description
Allocate this number of buffer pool buffers above the bar (ATB).
- Prior to version 7.7 of Model 204:
NUMBUFG specifies the number of buffer pool buffers built ATB, while the number of buffer pool buffers built BTB is based on the setting of MINBUF, MAXBUF, and SPCORE.
- As of version 7.7:
If NUMBUFG is specified, no buffer pool buffers are built BTB; all buffers are ATB. Model 204 obtains exactly the NUMBUFG number of buffers.
In this case, NUMBUFG essentially replaces the BTB parameters: NUMBUF and MINBUF are automatically reset to 0, and MAXBUF and SPCORE are ignored.
Setting NUMBUFG
The maximum number of buffers is 2**31/3, or 715,827,882, which limits buffer pools to 4.2 terabytes or less.
The minimum number of ATB buffers is:
NLRUQG * ((NSERVS + NSUBTKS) * MAXOBUF + 15)
If you set NUMBUFG to a value lower than the minimum, the value is reset to the calculated value and you receive messages like the following:
*** M204.2756: NUMBUFG reset to NLRUQG * ((NSERVS + NSUBTKS) * MAXOBUF + 15) = 1 * ((2+9) * 4 + 15) = 59 *** M204.0075: Initialization completed. Buffers = 0+59
Note:
If you set NUMBUFG greater than 0:
- The X'02' bit of the XMEMOPT parameter (the default), must also be set.
- The FLOD and FILELOAD commands and Fast/Reload are not available due to 64-bit addressing conflicts. This particular restriction is removed in V7.7 and later releases.
See also
For more information about setting NUMBUFG, see Using 64-bit addressing and Above the Bar (ATB) storage.