skmprabhu252@gmail.com has uploaded this change for review.
Config: Fix Long Parsing Time for Large Exports
The parser is experiencing prolonged processing times when handling
a large number of exports. This delay is primarily attributed to the
repetitive use of the strcmp function within the save_token() function.
To fix this, token string formatting was improved and hashing was
implemented for faster string searches, along with strlen comparison.
These changes led to a remarkable 95% boost in config parsing performance.
Without fix:
config_pointer = 0x8288f0config_pointer = 0x82cb60
real 28m38.646s
user 28m27.089s
With Fix:
config_pointer = 0x9a68f0config_pointer = 0x9aab70
real 1m34.930s
user 1m30.249s
sys 0m0.119s
Change-Id: I4084b1953e29f2d06abad9e639f9eadb1a133ca6
Signed-off-by: Prabhu Murugesan <prabhu.murugesan1@ibm.com>
---
M src/config_parsing/analyse.c
M src/config_parsing/analyse.h
2 files changed, 86 insertions(+), 50 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/23/1183423/1
To view, visit change 1183423. To unsubscribe, or for help writing mail filters, visit settings.