Re: [9fans] yet another try to fixup venti

[wb.kloke@gmail.com]

[-- Attachment #1: Type: text/plain, Size: 1372 bytes --]

I updated http://pkeus.de/~wb/mventi, adding a file mventi.c which is venti.c + parts of index.c (to avoid linking the real index.o).

I have some performance data now.

I had to add a 4th arenapartition, which sealed the old partitions, so readonly acces is sufficient for them,.
I served the partitions on the original venti system via nfs.

On my AMD A10 workstation (32MB, 4 cores)  it took  12min real time to vcat a 50GB ffs partiition to /dev/null.
On the WD Mycloud system ( on which the arena partitions are local, ithe same task took 33min.

mventi:
> time vcat ffs:ef6c183eabb7caafed7f85a3447a12accfbbeb8b > /dev/null
> 1540096 blocks total
> 1214607 blocks in use, 2754705 file reads
> 
> real  12m11,573s
> user  2m33,463s
> sys   0m34,604s

original venti:
> ime vcat ffs:ef6c183eabb7caafed7f85a3447a12accfbbeb8b > /dev/null
> 1540096 blocks total
> 1214607 blocks in use, 2754705 file reads
> 504.283u 94.546s 2000.095r     vcat ffs:ef6c183eabb7caafed7f85a3447a12accfbbeb8b
> 
I should add the remark, that I  was quite content with the performance of  the same  file served as a dump file system

------------------------------------------
9fans: 9fans
Permalink: https://9fans.topicbox.com/groups/9fans/T21878aa53884911b-M019008e416672a06797b66a8
Delivery options: https://9fans.topicbox.com/groups/9fans/subscription

[-- Attachment #2: Type: text/html, Size: 2365 bytes --]