Skip to main content

The server, me and the conversation

We were moving a project from AWS to our co-located DC. We have setup KVMs scheduled by Cloudstack for each of the component in the architecture. The KVMs used local storage. The VMs are provisioned with more than required resources because we have the opinion that in our DC scaling during peak load and then downscaling doesn't offer much benefits financially as we are anyways paying for the hardware in advance and its also powered on. Its going to be idle if not used. Now we found something interesting our latency in co-located DC was 2 times more than in AWS. The time for first byte at our load balancer in aws was 60ms average and at our DC was 112ms. We started our debugging mission, Mission Conquer-AWS. All the servers are newer Dell hardwares. So the initially intuition was virtualisation is causing the issue.

Conversation with the Hypervisor

We started with CPU optimisation, we started using the host-passthrough mode of CPU in libvirt so VMs dont see QEMU emulated CPUs, they see the actual CPU so they understand the flags supported by the underlying hardware and can use instruction set accordingly. We did VCPU pinning so that latency due to NUMA is ruled out (we know this should cause nano second latency not in millisecond). We stopped Kernel shared pages completely so that CPU time is not spent in finding identical pages and use Copy on Write if shared pages are written to. We also disabled transparent huge pages to avoid CPU time being spent on defragmenting memory for allocating huge pages. We didn't see any of the above optimisations yield noticeable impact in latency.

During the debugging we noted libvirt is configured to use ide for disks and e1000 for network calls. We understood KVM is running full virtualisation with HVM instead of Paravirtualised HVM. This causes KVM to do VMEXIT on an interrupt, handle it and do VMRUN. We restarted all vms with Virtio driver so the PV mode is on. On an interrupt guest via PV driver will talk to emulated QEMU device and do the IO. This reduced our latency to 100ms



















We are out of optimisations at hypervisor level and we are still 40ms behind AWS. There are still options available but all of them will improve in nano second which is not going to help us when we are over-provisioned(here over provisioned means provisioned beyond requirement, not to be confused with CPU overprovisioning terminology in VM world)

The next obvious choice was to test at baremetal directly and see how much it is giving raw performance. The baremetal gave pretty much same latency. This was something a shocker to us as unless AWS has a magic sauce of their own, beating baremetal with a virtualisation technology with such a huge margin is unimaginable. We removed hypervisor from the possible suspect list.

Conversation with the server

We saw the ping latency to VM was sometimes higher than the ping latency to bare metal directly. This took us to the following 
response by Greg https://www.spinics.net/lists/kvm/msg138556.html
The IO thread of KVM may be scheduled at a processor in C6 state and the latency may be due to wake up latency of the processor to C0 state. So we made IOthread busy with iperf and tried ping, the responsiveness of ping increased and we saw an improvement of 0.2 ms in ping response. We decided to fix the C state to 0 at the grub so that CPU always runs hot. This needed reboot of the whole infra as host machines had to be restarted with desired max C state. During this literature survey we also hit at P state. C state controls the sleep cycle of processors for power efficiency. P states control the frequency of the processors for power efficiency. Intel supports two scaling governors perfomance and powersave. Powersave uses P states which are biased toward powersaving efficiency than performance and this is the default P state in Linux. /proc/cpuinfo showed the CPUs running at 1.2GHz which is 50% of their 2.4GHz. So the instruction cycle runs double the time than it can run when hot. We switched the governor to performance mode. (https://wiki.archlinux.org/index.php/CPU_frequency_scaling)
Performance mode selects P states and set frequency accordingly where performance is biased over power efficiency. The performance improved and the latency came down to 60ms.
In the hindsight this looks like the first thing to be checked. 

Overall the mission is accomplished and we have started migrating traffic from AWS. Lesson Learnt - Look at the /proc/cpuinfo cpu MHz before getting into further optimisation

Comments

Popular posts from this blog

Lessons from Memory

Started debugging an issue where Linux started calling OOM reaper despite tons of memory is used as Linux cached pages. My assumption was if there is a memory pressure, cache should shrink and leave way for the application to use. This is the documented and expected behavior. OOM reaper is called when few number of times page allocation has failed consequently. If for example mysql wants to grow its buffer and it asks for a page allocation and if the page allocation fails repeatedly, kernel invokes oom reaper. OOM reaper won't move out pages, it sleeps for some time and sees if kswapd or a program has freed up caches/application pages. If not it will start doing the dirty job of killing applications and freeing up memory. In our mysql setup, mysql is the application using most of the Used Memory, so no other application can free up memory for mysql to use. Cached pages are stored as 2 lists in Linux kernel viz active and inactive.
More details here
https://www.kernel.org/doc/gorman…

How we have systematically improved the roads our packets travel to help data imports and exports flourish

This blog post is an account of how we have toiled over the years to improve the throughput of our interDC tunnels. I joined this company around 2012. We were scaling aggressively then. We quickly expanded to 4 DCs with a mixture of AWS and colocation. Our primary DC is connected to all these new DCs via IPSEC tunnels established from SRX. The SRX model we had, had an IPSEC throughput of 350Mbps. Around December 2015 we saturated the SRX. Buying SRX was an option on the table. Buying one with 2Gbps throughput would have cut the story short. The tech team didn't see it happening.

I don't have an answer to the question, "Is it worth spending time in solving a problem if a solution is already available out of box?" This project helped us in improving our critical thinking and in experiencing the theoretical network fundamentals on live traffic, but also caused us quite a bit of fatigue due to management overhead. Cutting short the philosophy, lets jump to the story.

De…

Walking down the Memory Lane!!!

This post is going to be an account of  few trouble-shootings I did recently to combat various I/O sluggishness.
Slow system during problems with backup
We have a NFS mount where we push backups of our database daily. Due to some update to the NFS infra, we started seeing throughput of NFS server drastically affected. During this time we saw general sluggishness in the system during backups. Even ssh logins appeared slower. Some boxes had to be rebooted due to this sluggishness as they were too slow to operate on them. First question we wanted to answer, does NFS keep writing if the server is slow? The slow server applied back pressure by sending small advertised window(TCP) to clients. So clients can't push huge writes if server is affected. Client writes to its page cache. The data from page cache is pushed to server when there is a memory pressure or file close is called. If server is slow, client can easily reach upto dirty_background_ratio set for page cache in sysctl. This di…