blob: 2a278c6361c40d681dfcf8d957ec5af15549c9d9 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
|
I'm running obnam (1st run) on a decent sized (2.8T) volume. It's been running for days and beyond being slow, it's using a good bit of memory. This does not seem reasonable. Is there something I should tweak?
Thanks
(unloaded switched gigE between server/client)
<pre>
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
20848 root 20 0 1745m 1.4g 1620 R 70 68.9 8558:22 obnam
# more .obnam.conf
[config]
repository: sftp://xxx/yyy/obnam/
log-level: warning
log: /root/obnam.log
log-max: 10M
weak-random: true
lock-timeout: 10
compress-with: deflate
</pre>
---
I have not benchmarked Obnam's memory usage for a while, so I don't know if that is unexpectedly large.
It's certainly larger than I would like, though. You can try different sizes of the following settings:
* `node-size`
* `upload-queue-size`
* `lru-size`
If you change the node size, you'll need to remove the repository and start over, since it only affects new
instances of the in-repository data structures, so it's probably best to start by trying different values for the
other two settings.
--liw
--
I've reduced the default values for lru-size and upload-queue-size. Obnam should now be
using a lot less memory. --liw
[[done]]
|
