Github-Mirror/seaweedfs
1
0
Fork 0
mirror of https://github.com/seaweedfs/seaweedfs.git synced 2026-09-20 13:30:46 +02:00
Code Issues Packages Projects Releases Wiki Activity
826 Commits 181 Branches 360 Tags
3a024d62df820fa5f932e331c0c49e1e09fb6ed2
Commit Graph
21 Commits
Author SHA1 Message Date
chrislusf 1b6ab2f6af Add boltdb for volume needle map 
boltdb is fairly slow to write, about 6 minutes for recreating index
for 1553934 files. Boltdb loads 1,553,934 x 16 = 24,862,944bytes from
disk, and generate the boltdb as large as 134,217,728 bytes in 6
minutes.

To compare, for leveldb, it recreates index in leveldb as large as
27,188,148 bytes in 8 seconds.
For in memory version, it loads the index in

To test the memory consumption, the leveldb or boltdb index are
created. And the server is restarted. Using the benchmark tool to read
lots of files. There are 7 volumes in benchmark collection, each with
about 1553K files.
For leveldb, the memory starts at 142,884KB, and stays at 179,340KB.
For boltdb, the memory starts at 73,756KB, and stays at 144,564KB.
For in-memory, the memory starts at 368,152KB, and stays at 448,032KB.
 2015-03-29 11:04:32 -07:00
chrislusf 020ba6c9a8 add leveldb support for needle map 
This supposedly should reduce memory consumption. However, for tests
with millions of, this shows consuming more memories. Need to see
whether this will work out. If not, later boltdb will be tested.
 2015-03-27 16:34:58 -07:00
chrislusf 9d8a6d2562 adjust log printing 2015-03-09 01:10:04 -07:00
Chris Lu af416189f1 Cleanup error printing. 2015-01-13 17:04:41 -08:00
Chris Lu 179d36ba0e formatting code by: goimports -w=true . 2014-10-26 11:34:55 -07:00
wyy 4126280d55 use github.com/chrislusf instead of github.com/aszxqw 2014-09-25 16:57:22 +08:00
wyy 1cd19447e3 use github.com/aszxqw instead of code.google.com/p 2014-09-25 00:47:09 +08:00
Chris Lu faf497feb8 Important Fix: Avoid overwriting the index file!!! 2014-05-29 19:19:03 -07:00
Chris Lu 3b5035c468 1. v0.54 
2. go vet found many printing format errors
 2014-04-17 00:16:44 -07:00
Chris Lu 51939efeac 1. volume server now sends master server its max file key, so that 
master server does not need to store the sequence on disk any more
2. fix raft server's failure to init cluster during bootstrapping
 2014-04-16 23:43:27 -07:00
Chris Lu 3dbebfd1e1 Thread-safe fixes: 
1. avoid sharing []byte
2. switch to use ReadAt()
 2014-03-19 04:41:16 -07:00
Chris Lu cd10c277b2 can now delete a collection! Is this a dangerous feature? Only enabling 
deleting "benchmark" collections for now.
 2014-03-10 11:43:54 -07:00
Chris Lu 82b74c7940 issue 43 "go fmt" chagnes from "Ryan S. Brown" <sb@ryansb.com> 
some basic changes to parse upload url
 2013-09-01 23:58:21 -07:00
Chris Lu ed154053c8 switching to temporarily use glog library 2013-08-08 23:57:22 -07:00
Chris Lu 50269b74ce add dataCenter option when assign file keys 
add dataCenter option when starting volume servers
some work related to freeze a volume. Not tested yet.
 2013-06-19 18:10:38 -07:00
Chris Lu fb635146a1 refactoring needle mapper interface to separate index file storage logic 
out
 2013-04-18 00:23:14 -07:00
Chris Lu db8e27be6e add lots of error checking by GThomas 2013-02-26 22:54:22 -08:00
Chris Lu a4cef2fbd4 only cares about error other than EOF 2013-02-11 23:54:21 -08:00
Chris Lu 91492b4947 avoid EOF error on windows. This is a "try to load" function anyway. 2013-02-11 23:38:48 -08:00
Chris Lu aecea22666 Issue 15: Stress test corrupts volume 2013-02-10 14:00:06 -08:00
Chris Lu 5071f528f6 testing compilation with remove package 2013-02-10 03:49:51 -08:00