Novell EDIRECTORY 8.8 SP3, Administration Manual, Page: 537 / 634

Maintaining Novell eDirectory
18
537
n
ov
do
cx (e
n)
11
Ju
ly 20
08
18
Maintaining Novell eDirectory
For Novell
®
eDirectory
TM
to perform optimally, you need to maintain the directory through routine
health check procedures and upgrading or replacing hardware when necessary.
This chapter covers the following maintenance topics:
Performance
Š
Section 18.1, “Improving eDirectory Performance,” on page 537
Š Section 18.2, “Improving eDirectory Performance on Linux, Solaris, and AIX Systems,” on
page 545
Š
Section 18.4, “Advanced Referral Costing,” on page 551
Š
Section 18.5, “Improving Bulkload Performance,” on page 560
Š
Section 18.6, “Countering Memory Fragmentation,” on page 565
Health Checks
Š
Section 18.7, “Keeping eDirectory Healthy,” on page 566
Š
Section 18.8, “Resources for Monitoring,” on page 569
Hardware Replacements
Š
Section 18.9, “Upgrading Hardware or Replacing a Server,” on page 569
eDirectory Recovery
Š
Section 18.10, “Restoring eDirectory after a Hardware Failure,” on page 576
18.1 Improving eDirectory Performance
The most significant setting that affects eDirectory performance is the cache. In earlier versions of
NDS
®
, you could specify a block cache limit to regulate the amount of memory that the directory
used for the cache. The default was 8 MB RAM for cache.
With eDirectory 8.5 or later, you can specify a block cache limit and an entry cache limit. The block
cache, available in earlier versions of NDS, caches only physical blocks from the database. The
entry cache, a feature introduced in eDirectory 8.5, caches logical entries from the database. The
caching of entries reduces the processing time required to instantiate entries in memory from the
block cache.
Although there is some redundancy between the two caches, each cache is designed to boost
performance for different operations. Block cache is most useful for update operations. Entry cache
is most useful for operations that browse the eDirectory tree by reading through entries, such as
name resolution.
Both block and entry caches are useful in improving query performance. Block cache speeds up
index searching. Entry cache speeds up the retrieval of entries referenced from an index.