Glad you have a root cause/solution. On Sat, Jan 8, 2011 at 10:49 AM, Simon <[email protected]> wrote: <SNIP> > The virtual HD is physically on a raid (unknown config). Mark, the sector > size issue you mention, does it have to do with aligning real HD sectors > with filesystem sectors (so that stuff like read-ahead will get > no-more-no-less than what the kernel wants)? I've read about this kind of > setup when I was interested in RAID long ago... Now that I know my hd is > actually on a raid, maybe i could benefit some I/O performance improvements > by tuning this a bit! >
As it's RAID underneath it's likely set up correctly. The issue I had in mind was the disk being a 4K/sector disk but the person who built the partition not knowing to align the partition to a 4K boundary. That can cause a _huge_ slowdown. I doubt that's the case here. As this is a hosting service they likely know what they are doing in that area, and if it wasn't done correctly you would have noticed it before I think. > Anyway, I was told by the support team that another user on the same > physical machine (remember it's a xen VPS) was doing I/O intensive stuff > which could have "I/O starved" my system. I don't understand how starving > or even doing some kind of DoS attack could lead to a complete freeze on the > console, but eh... Makes sense actually. The other guy took all the disk I/O leaving you with none. If you can't get to the disk then you cannot read ebuilds or write compiled code, or at least not fast. > They offered to migrate my system to another physical > machine, and after that... I was able to perform a complete 'emerge -e > system' in one shot without a scratch, I even did it with --jobs=2 and > MAKEOPTS="-j4". After that, I started a complete "emerge --keep-going > --jobs=2 world" with MAKEOPTS="-j8"... (i got 4 cores: dual xeon 2Ghz) > So now you're in good shape...until some user on the new system starts hogging all the disk I/O and holds you up again. > This last emerge is still going on as I write this and is emerging pkg 522 > of 620 !! And there were no build errors so far... > > It's emerging glibc at the moment, so once the big emerge is finished, I'll > probably recompile all pkgs that depend on glibc. I believe glibc was > actually updated during my very initial update on monday and I haven't come > to do that... but I guess everything will go smoothly from here. > > Thanks again for all your help guys! > Simon Good that you got to the root of the problem. Good luck, Mark
