Message ID | 20230202123530.945140-1-i.maximets@ovn.org |
---|---|
Headers | show |
Series | northd: Hash locks for dp groups + thread safety. | expand |
On 2/2/23 13:35, Ilya Maximets wrote: > While running tests with ovn-heater it was observed that locking and > unlocking dpg_lock can take more than 20% of CPU cycles in northd even > without any contention. > > This series is trying to address that issue and add thread safety > static analysis. > Hi Ilya, Thanks for these patches, this looks like a very nice performance boost indeed! Before doing a proper review, would it be possible to rebase them on top of the current main branch? I just pushed Ales' '7b94d212f694 ("northd: Refactor build_lrouter_nat_flows_for_lb function")' v6 and that created some conflicts in northd.c Thanks, Dumitru
On 2/6/23 12:49, Dumitru Ceara wrote: > On 2/2/23 13:35, Ilya Maximets wrote: >> While running tests with ovn-heater it was observed that locking and >> unlocking dpg_lock can take more than 20% of CPU cycles in northd even >> without any contention. >> >> This series is trying to address that issue and add thread safety >> static analysis. >> > > Hi Ilya, > > Thanks for these patches, this looks like a very nice performance boost > indeed! > > Before doing a proper review, would it be possible to rebase them on top > of the current main branch? I just pushed Ales' '7b94d212f694 ("northd: > Refactor build_lrouter_nat_flows_for_lb function")' v6 and that created > some conflicts in northd.c Done. Though I've gone a bit wild with the set. :) The conflict wasn't an easy one to resolve. Best regards, Ilya Maximets.
On 2/7/23 12:45, Ilya Maximets wrote: > On 2/6/23 12:49, Dumitru Ceara wrote: >> On 2/2/23 13:35, Ilya Maximets wrote: >>> While running tests with ovn-heater it was observed that locking and >>> unlocking dpg_lock can take more than 20% of CPU cycles in northd even >>> without any contention. >>> >>> This series is trying to address that issue and add thread safety >>> static analysis. >>> >> >> Hi Ilya, >> >> Thanks for these patches, this looks like a very nice performance boost >> indeed! >> >> Before doing a proper review, would it be possible to rebase them on top >> of the current main branch? I just pushed Ales' '7b94d212f694 ("northd: >> Refactor build_lrouter_nat_flows_for_lb function")' v6 and that created >> some conflicts in northd.c > > Done. Though I've gone a bit wild with the set. :) The conflict wasn't > an easy one to resolve. Sorry about that and thanks for taking care of it! I'll try to review v2 as soon as possible. Regards, Dumitru