On Thu 19-04-12 22:23:43, Wu Fengguang wrote: > For one instance, splitting the request queues will give rise to > PG_writeback pages. Those pages have been the biggest source of > latency issues in the various parts of the system. Well, if we allow more requests to be in flight in total then yes, number of PG_Writeback pages can be higher as well. > It's not uncommon for me to see filesystems sleep on PG_writeback > pages during heavy writeback, within some lock or transaction, which in > turn stall many tasks that try to do IO or merely dirty some page in > memory. Random writes are especially susceptible to such stalls. The > stable page feature also vastly increase the chances of stalls by > locking the writeback pages. > > Page reclaim may also block on PG_writeback and/or PG_dirty pages. In > the case of direct reclaim, it means blocking random tasks that are > allocating memory in the system. > > PG_writeback pages are much worse than PG_dirty pages in that they are > not movable. This makes a big difference for high-order page allocations. > To make room for a 2MB huge page, vmscan has the option to migrate > PG_dirty pages, but for PG_writeback it has no better choices than to > wait for IO completion. > > The difficulty of THP allocation goes up *exponentially* with the > number of PG_writeback pages. Assume PG_writeback pages are randomly > distributed in the physical memory space. Then we have formula > > P(reclaimable for THP) = 1 - P(hit PG_writeback)^256 Well, this implicitely assumes that PG_Writeback pages are scattered across memory uniformly at random. I'm not sure to which extent this is true... Also as a nitpick, this isn't really an exponential growth since the exponent is fixed (256 - actually it should be 512, right?). It's just a polynomial with a big exponent. But sure, growth in number of PG_Writeback pages will cause relatively steep drop in the number of available huge pages. ... > It's worth to note that running multiple flusher threads per bdi means > not only disk seeks for spin disks, smaller IO size for SSD, but also > lock contentions and cache bouncing for metadata heavy workloads and > fast storage. Well, this heavily depends on particular implementation (and chosen data structures). But yes, we should have that in mind. ... > > > To me, balance_dirty_pages() is *the* proper layer for buffered writes. > > > It's always there doing 1:1 proportional throttling. Then you try to > > > kick in to add *double* throttling in block/cfq layer. Now the low > > > layer may enforce 10:1 throttling and push balance_dirty_pages() away > > > from its balanced state, leading to large fluctuations and program > > > stalls. > > > > Just do the same 1:1 inside each cgroup. > > Sure. But the ratio mismatch I'm talking about is inter-cgroup. > For example there are only 2 dd tasks doing buffered writes in the > system. Now consider the mismatch that cfq is dispatching their IO > requests at 10:1 weights, while balance_dirty_pages() is throttling > the dd tasks at 1:1 equal split because it's not aware of the cgroup > weights. > > What will happen in the end? The 1:1 ratio imposed by > balance_dirty_pages() will take effect and the dd tasks will progress > at the same pace. The cfq weights will be defeated because the async > queue for the second dd (and cgroup) constantly runs empty. Yup. This just shows that you have to have per-cgroup dirty limits. Once you have those, things start working again. Honza -- Jan Kara <jack@xxxxxxx> SUSE Labs, CR -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html