Clustered Samba: Difference between revisions
(link to svg sources) |
m (typo) |
||
Line 1: | Line 1: | ||
Current [[Samba]] implementations can't be used directly on [[Cluster|clustered systems]] cause |
Current [[Samba]] implementations can't be used directly on [[Cluster|clustered systems]] cause they are oriented to single-server systems. The main reason is that [[smbd]] processes use [[TDB]] (local-oriented database) for [[messaging]], storing shared data, etc - there is no way to coordinate [[smbd]] processes run on different cluster nodes. |
||
Simple [[clustered samba]] system looks like this: |
Simple [[clustered samba]] system looks like this: |
Revision as of 09:48, 10 May 2006
Current Samba implementations can't be used directly on clustered systems cause they are oriented to single-server systems. The main reason is that smbd processes use TDB (local-oriented database) for messaging, storing shared data, etc - there is no way to coordinate smbd processes run on different cluster nodes.
Simple clustered samba system looks like this:
(you can get all images's SVG sources here - [1])
Each node has its own smbd daemon - they should communicate with each other to avoid shared data corruption, treat oplocks and so on. So the most problem is the extending of the locking subsystem (share mode locks, opportunistic locks, byte-range locks) to multi-node system and cluster point of view.
The Locking Problem
Here we'll gather information on how locking mechanisms are done in Samba3.
Present locking mechanism uses single locking.tdb (database with shared information - share mode entries, deferred open entries, flags, etc) for all smbd processes:
Course of development
On clustered systems internal messaging can be used to interconnect smbd processes and locking databases:
This approach implies too many messaging opearations between smbd processes while locking/unlocking a file. They can be reduced noticeably if all locking information is stored on the one node - with the locking daemon: