Re: FSFS nosync

2011-12-16 Thread Peter Samuelson
[Stefan Sperling] > On Thu, Dec 15, 2011 at 04:04:13PM -0600, Peter Samuelson wrote: > > http://packages.debian.org/sid/eatmydata > > https://launchpad.net/libeatmydata > > > > It apparently works on Linux and Solaris. Don't know if that's enough > > coverage for general interest. > Eve

AW: FSFS nosync

2011-12-16 Thread Markus Schaber
Hi, Von: Branko Čibej [mailto:br...@xbc.nu] >On 15.12.2011 23:38, Stefan Sperling wrote: >> On Thu, Dec 15, 2011 at 04:04:13PM -0600, Peter Samuelson wrote: >>> [Philip Martin] If we had such a flag in fsfs.conf (Stefan suggests "eat-my-data=yes") the code could write all the same data

Re: FSFS nosync

2011-12-16 Thread Branko Čibej
On 15.12.2011 23:38, Stefan Sperling wrote: > On Thu, Dec 15, 2011 at 04:04:13PM -0600, Peter Samuelson wrote: >> [Philip Martin] >>> If we had such a flag in fsfs.conf (Stefan suggests >>> "eat-my-data=yes") the code could write all the same data in the same >>> order but avoid making any flush ca

Re: FSFS nosync

2011-12-15 Thread Stefan Sperling
On Thu, Dec 15, 2011 at 04:04:13PM -0600, Peter Samuelson wrote: > > [Philip Martin] > > If we had such a flag in fsfs.conf (Stefan suggests > > "eat-my-data=yes") the code could write all the same data in the same > > order but avoid making any flush calls thus allowing the OS to order > > physic

Re: FSFS nosync

2011-12-15 Thread Peter Samuelson
[Philip Martin] > If we had such a flag in fsfs.conf (Stefan suggests > "eat-my-data=yes") the code could write all the same data in the same > order but avoid making any flush calls thus allowing the OS to order > physical writes for optimum speed. Given the main use case is a distinct svnadmin

Re: FSFS nosync

2011-12-15 Thread Greg Stein
On Dec 15, 2011 1:26 PM, "Stefan Sperling" wrote: > > On Thu, Dec 15, 2011 at 01:04:04PM -0500, Greg Stein wrote: > > Couldn't we just make that an option for loading, but not provide such a > > feature for normal operation? That seems safer to me, and solves the actual > > use case. > > That woul

Re: FSFS nosync

2011-12-15 Thread Stefan Sperling
On Thu, Dec 15, 2011 at 01:04:04PM -0500, Greg Stein wrote: > Couldn't we just make that an option for loading, but not provide such a > feature for normal operation? That seems safer to me, and solves the actual > use case. That would require revving the repos and fs APIs. Hence the idea of putti

Re: FSFS nosync

2011-12-15 Thread Greg Stein
Couldn't we just make that an option for loading, but not provide such a feature for normal operation? That seems safer to me, and solves the actual use case. +1 on the flag name :-) Cheers, -g On Dec 15, 2011 10:59 AM, "Philip Martin" wrote: > From a discussion on IRC: > > A BDB repository all

FSFS nosync

2011-12-15 Thread Philip Martin
>From a discussion on IRC: A BDB repository allows the admin to set DB_TXN_NOSYNC in the DBD configuration file, this allows the admin to trade performance for robustness. We could do something similar in FSFS. When loading a dumpfile into a FSFS repository I see 13 calls to fsync per revision o