| 1 |
"John R. Dunning" writes: |
| 2 |
> |
| 3 |
> Lustre 1.6 (at least the client end) doesn't even really *require* all those |
| 4 |
> kernel patches, ie they do support the idea of a patchless client. The issue |
| 5 |
> is that lustre changes the logic involved in various kinds of fs operations, |
| 6 |
> including anything related to lookups, so as to short-circuit much of the wor |
| 7 |
> k |
| 8 |
> involved when it figures out that it can do so. Running the client without |
| 9 |
> the patches will work, but it won't give you the performance that you'd get |
| 10 |
> with the patches. So odds are anybody who's interested in running lustre in |
| 11 |
> the first place probably wants the patches too. |
| 12 |
|
| 13 |
I hadn't realized that the patchless client was potentially |
| 14 |
lower-performance than a patched client. Are you sure about that? |
| 15 |
How much of a difference do you think it is? |
| 16 |
Are you using version 1.6 or 1.4? |
| 17 |
|
| 18 |
> |
| 19 |
> We at sicortex are planning on rolling out a gentoo-based cluster that depend |
| 20 |
> s |
| 21 |
> heavily on lustre, so we've spent a fair bit of time banging on it. I'm |
| 22 |
> pretty sure we understand it at this point. We'll know for sure soon :-} |
| 23 |
|
| 24 |
Do you get support from CFS? It seems pretty clear that you do not. |
| 25 |
What kernel versions do you use? |
| 26 |
|
| 27 |
-bryan |
| 28 |
|
| 29 |
-- |
| 30 |
gentoo-cluster@g.o mailing list |
Archives