| To: | Steve Lord <lord@xxxxxxx> |
|---|---|
| Subject: | Re: synchronization of XFS |
| From: | Stefan Smietanowski <stesmi@xxxxxxxxxx> |
| Date: | Thu, 08 Apr 2004 03:32:35 +0200 |
| Cc: | "IKARASHI, Seiichi" <ikarashi@xxxxxxxxxxxxxxxx>, Christoph Hellwig <hch@xxxxxxxxxxxxx>, Chris Wedgwood <cw@xxxxxxxx>, linux-xfs@xxxxxxxxxxx |
| In-reply-to: | <40682301.50508@xfs.org> |
| References: | <4060F7FC.8090602@miraclelinux.com> <20040325063902.GA9697@dingdong.cryptoapps.com> <4062C97A.6030702@miraclelinux.com> <20040325124152.GA12078@dingdong.cryptoapps.com> <4062D7E5.6070501@stesmi.com> <20040325132200.GA12333@dingdong.cryptoapps.com> <4062E19A.90207@xfs.org> <20040325140723.GA12558@dingdong.cryptoapps.com> <20040325144519.A23764@infradead.org> <40635F04.6010109@xfs.org> <40636032.3000402@stesmi.com> <4063612E.4030109@xfs.org> <406361F2.6060308@stesmi.com> <4063650B.20600@xfs.org> <406365DE.2050006@stesmi.com> <4067B496.8060501@miraclelinux.com> <40682301.50508@xfs.org> |
| Sender: | linux-xfs-bounce@xxxxxxxxxxx |
| User-agent: | Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7b) Gecko/20040316 |
Steve Lord wrote: > IKARASHI, Seiichi wrote: > >> >> How was this testing result? >> I suppose setting sync_interval to 1000 and sleeping 2 seconds >> after sync is basically same as sleeping 4 seconds or more with >> the default sync_interval value (3000?). >> >> Seiich >> > Except the default setting is 30000 or 30 seconds. > > I have not heard back from Stefan. > > Steve Alright. I took my x86-64 machine and installed FC2-test2 on it and it has the problem (original). I then just took the DVD and burnt me a new one. Same problem, naturally. Then I changed booty and added some code that checked if /proc/sys/fs/xfs/sync_interval existed and was writable at all and if it was, it wrote 1000 to it (1 second). Then I added a time.sleep(5) (5 seconds delay), a few more isys.sync() after that and finally another time.sleep(5) just to be on the safe side. And guess what - still no go :( At least I now have something I can work with again. Any more ideas? The kernel is based on 2.6.3 and I haven't looked into it if they patch anything relevant for us. Any more things anyone wants me to try - go right ahead and ask - it now takes me only roughly an hour to build and burn and test it, so it's not really painful as most of it's automated. // Stefan |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | TAKE 912498 - fix 2.6 page state corruption, Nathan Scott |
|---|---|
| Next by Date: | Re: ST alloc failures, Tim Shimmin |
| Previous by Thread: | TAKE 912498 - fix 2.6 page state corruption, Nathan Scott |
| Next by Thread: | Re: synchronization of XFS, Chris Wedgwood |
| Indexes: | [Date] [Thread] [Top] [All Lists] |