FGBI

Traditional xen-based systems track memory updates by keeping evidence of the dirty pages at each migration epoch. In ​Remus and also our previous work, LLM, they use the same page size as Xen (for x86, this is 4KB), which is also the granularity for detecting memory changes. However, when running computational-intensive workloads under LLM system, the long downtime performance becomes unacceptable. FGBI (Fine-Grained Block Identification) is a mechanism which uses smaller memory blocks (smaller than page sizes) as the granularity for detecting memory changes. FGBI calculates the hash value for each memory block at the beginning of each migration epoch. At the end of each epoch, instead of transferring the whole dirty page, FGBI computes new hash values for each block and compares them with the corresponding old values. Blocks are only modified if their corresponding hash values don’t match. Therefore, FGBI marks such blocks as dirty and replaces the old hash values with the new ones. Afterwards, FGBI only transfers dirty blocks to the backup host.

FGBI is based on ​The Remus project and our previous efforts Lightweight Live Migration (LLM) mechanism. For a full description and evaluation, please see our OPODIS'11 paper.

Downtime Problem in LLM

Figure 1. Primary-Backup model and the downtime problem.

Downtime is the primary factor for estimating the high availability of a system, since any long downtime experience for clients may result in loss of client loyalty and thus revenue loss. Under the Primary-Backup model (Figure 1), there are two types of downtime: I) the time from when the primary host crashes until the VM resumes from the last checkpointed state on the backup host and starts to handle client requests (D1 = T3 - T1); II) the time from when the VM pauses on the primary (to save for the checkpoint) until it resumes (D2). From Jiang’s paper we observe that for memory-intensive workloads running on guest VMs (such as the highSys workload), LLM endures much longer type I downtime than Remus. This is because, these workloads update the guest memory at high frequency. On the other side, LLM migrates the guest VM image update (mostly from memory) at low frequency but uses input replay as an auxiliary. In this case, when failure happens, a significant number of memory updates are needed in order to ensure synchronization between the primary and backup hosts. Therefore, it needs significantly more time for the input replay process in order to resume the VM on the backup host and begin handling client requests.

Regarding the type II downtime, there are several migration epochs between two checkpoints, and the newly updated memory data is copied to the backup host at each epoch. At the last epoch, the VM running on the primary host is suspended and the remaining memory states are transferred to the backup host. Thus, the type II downtime depends on the amount of memory that remains to be copied and transferred when pausing the VM on the primary host. If we reduce the dirty data which need to be transferred at the last epoch, then we can reduce the type II downtime. Moreover, if we reduce the dirty data which needs to be transferred at each epoch, trying to synchronize the memory state between primary and backup host all the time, then at the last epoch, there won’t be too much new memory update that need to be transferred, so we can reduce the type I downtime too.

FGBI Design

Therefore, in order to achieve HA in these virtualized systems, especially to address the downtime problem under memory-intensive workloads, we propose a memory synchronization technique for tracking memory updates, called Fine-Grained Block Identification (or FGBI). Remus and LLM track memory updates by keeping evidence of the dirty pages at each migration epoch. Remus uses the same page size as Xen (for x86, this is 4KB), which is also the granularity for detecting memory changes. However, this mechanism is not efficient. For instance, no matter what changes an application makes to a memory page, even just modify a boolean variable, the whole page will still be marked dirty. Thus, instead of one byte, the whole page needs to be transferred at the end of each epoch. Therefore, it is logical to consider tracking the memory update at a finer granularity, like dividing the memory into smaller blocks.

We propose the FGBI mechanism which uses memory blocks (smaller than page sizes) as the granularity for detecting memory changes. FBGI calculates the hash value for each memory block at the beginning of each migration epoch. Then it uses the same mechanism as Remus to detect dirty pages. However, at the end of each epoch, instead of transferring the whole dirty page, FGBI computes new hash values for each block and compares them with the corresponding old values. Blocks are only modified if their corresponding hash values do not match. Therefore, FGBI marks such blocks as dirty and replaces the old hash values with the new ones. Afterwards, FGBI only transfers dirty blocks to the backup host. However, because of using block granularity, FGBI introduces new overhead. If we want to accurately approximate the true dirty region, we need to set the block size as small as possible. For example, to obtain the highest accuracy, the best block size is one bit. That is impractical because it requires storing an additional bit for each bit in memory, which means that we need to double the main memory. Thus, a smaller block size leads to a greater number of blocks and also requires more memory for storing the hash values. Based on these past e orts illustrating the memory saving potential, we present two supporting techniques: block sharing and hybrid compression.

Downtime Evaluations

Figures 2a, 2b, 2c, and 2d show the type I downtime com- parison among FGBI, LLM, and Remus mechanisms under Apache, NPB-EP, SPECweb, and SPECsys applications, respectively. The block size used in all experiments is 64 bytes. For Remus and FGBI, the checkpointing period is the time interval of system update migration, whereas for LLM, the checkpointing period represents the interval of network buffer migration. By configuring the same value for the checkpointing frequency of Remus/FGBI and the network buffer frequency of LLM, we ensure the fairness of the comparison. We observe that Figures 2a and 2b show a reverse relationship between FGBI and LLM. Under Apache (Figure 2a), the network load is high but system updates are rare. Therefore, LLM performs better than FGBI, since it uses a much higher frequency to migrate the network service requests. On the other hand, when running memory-intensive applications (Figure 2b and 2d), which involve high computational loads, LLM endures a much longer downtime than FGBI (even worse than Remus).

Although SPECweb is a web workload, it still has a high page modifi- cation rate, which is approximately 12,000 pages/second. In our experi- ment, the 1 Gbps migration link is capable of transferring approximately 25,000 pages/second. Thus, SPECweb is not a lightweight computational workload for these migration mechanisms. As a result, the relationship between FGBI and LLM in Figure 2c is more similar to that in Figure 2b (and also Figure 2d), rather than Figure 2a. In conclusion, compared with LLM, FGBI reduces the downtime by as much as 77%. Moreover, compared with Remus, FGBI yields a shorter downtime, by as much as 31% under Apache, 45% under NPB-EP, 39% under SPECweb, and 35% under SPECsys.

Table 1 shows the type II downtime comparison among Remus, LLM, and FGBI mechanisms under different applications. We have three main observations: (1) Their downtime results are very similar for "idle" run. This is because Remus is a fast checkpointing mechanism and both LLM and FGBI are based on it. There is rare memory update for "idle" run, so the type II downtime in all three mechanisms is short. (2) When running NPB-EP ap- plication, the guest VM memory is updated at high frequency. When saved for the checkpoint, LLM takes much more time to save huge "dirty" data caused by its low memory transfer frequency. Therefore in this case FGBI achieves a much lower downtime than Remus (reduce more than 70%) and LLM (more than 90%). (3) When running Apache application, the memory update is not so much as that when running NPB, but the memory update is definitely more than "idle" run. The downtime results shows FGBI still outperforms both Remus and LLM.

Overhead

Figure 3a shows the overhead during VM migration. The figure compares the applications' runtime with and without migration, under Apache, SPECweb, NPB-EP, and SPECsys, with the size of the fine-grained blocks varies from 64 bytes to 128 bytes and 256 bytes. We observe that in all cases the overhead is low, no more than 13% (Apache with 64 bytes block). As we discuss in Section 3, the smaller the block size that FGBI chooses, the greater is the memory overhead that it introduces. In our experiments, the smaller block size that we chose is 64 bytes, so this is the worst case overhead compared with the other block sizes. Even in this "worst" case, under all these benchmarks, the overhead is less than 8.21%, on average.

In order to understand the respective contributions of the three proposed techniques (i.e., FGBI, sharing, and compression), Figure 3b shows the break- down of the performance improvement among them under the NPB-EP bench- mark. It compares the downtime between integrated FGBI (which we use for evaluation in this Section), FGBI with sharing but no compression support, FGBI with compression but no sharing support, and FGBI without sharing nor compression support, under the NPB-EP benchmark. As previously discussed, since NPB-EP is a memory-intensive workload, it should present a clear differ- ence among the three techniques, all of which focus on reducing the memory- related overhead. We do not include the downtime of LLM here, since for this compute-intensive benchmark, LLM incurs a very long downtime, which is more than 10 times the downtime that FGBI incurs.

We observe from Figure 3b that if we just apply the FGBI mechanism without integrating sharing or compression support, the downtime is reduced, compared with that of Remus in Figure 4b, but it is not significant (reduction is no more than twenty percent). However, compared with FGBI with no support, after in- tegrating hybrid compression, FGBI further reduces the downtime, by as much as 22%. We also obtain a similar benefit after adding the sharing support (down- time reduction is a further 26%). If we integrate both sharing and compression support, the downtime is reduced by as much as 33%, compared to FGBI without sharing or compression support.