|| ||Fernando Luis Vazquez Cao <firstname.lastname@example.org> |
|| ||email@example.com |
|| ||[RFC] High availability in KVM |
|| ||Thu, 17 Jun 2010 12:15:20 +0900|
|| ||Linux Kernel Mailing List <firstname.lastname@example.org>,
Takuya Yoshikawa <email@example.com>,
Chris Wright <firstname.lastname@example.org>, Dor Laor <email@example.com>,
Lon Hohberger <firstname.lastname@example.org>,
"Perry N. Myers" <email@example.com>,
Luiz Capitulino <firstname.lastname@example.org>, email@example.com|
|| ||Article, Thread
We are trying to improve the integration of KVM with the most common
HA stacks, but we would like to share with the community what we are
trying to achieve and how before we take a wrong turn.
This is a pretty long write-up, but please bear with me.
Virtualization has boosted flexibility on the data center, allowing
for efficient usage of computer resources, increased server
consolidation, load balancing on a per-virtual machine basis -- you
name it, However we feel there is an aspect of virtualization that
has not been fully exploited so far: high availability (HA).
Traditional HA solutions can be classified in two groups: fault
tolerant servers, and software clustering.
Broadly speaking, fault tolerant servers protect us against hardware
failures and, generally, rely on redundant hardware (often
proprietary), and hardware failure detection to trigger fail-over.
On the other hand, software clustering, as its name indicates, takes
care of software failures and usually requires a standby server
whose software configuration for the part we are trying to make
fault tolerant must be identical to that of the active server.
Existing open source HA stacks such as pacemaker/corosync and Red
Hat Cluster Suite rely on software clustering techniques to detect
both hardware failures and software failures, and employ fencing to
avoid split-brain situations which, in turn, makes it possible to
perform failover safely. However, when applied to virtualization
environments these solutions show some limitations:
- Hardware detection relies on polling mechanisms (for example
pinging a network interface to check for network connectivity),
imposing a trade off between failover time and the cost of
polling. The alternative is having the failing system send an
alarm to the HA software to trigger failover. The latter
approach is preferable but it is not always applicable when
dealing with bare-metal; depending on the failure type the
hardware may not able to get a message out to notify the HA
software. However, when it comes to virtualization environments
we can certainly do better. If a hardware failure, be it real
hardware or virtual hardware, is fully contained within a
virtual machine the host or hypervisor can detect that and
notify the HA software safely using clean resources.
- In most cases, when a hardware failure is detected the state of
the failing node is not known which means that some kind of
fencing is needed to lock resources away from that
node. Depending on the hardware and the cluster configuration
fencing can be a pretty expensive operation that contributes to
system downtime. Virtualization can help here. Upon failure
detection the host or hypervisor could put the virtual machine
in a quiesced state and release its hardware resources before
notifying the HA software, so that it can start failover
immediately without having to mingle with the failing virtual
machine (we now know that it is in a known quiesced state). Of
course this only makes sense in the event-driven failover case
- Fencing operations commonly involve killing the virtual machine,
thus depriving us of potentially critical debugging information:
a dump of the virtual machine itself. This issue could be solved
by providing a virtual machine control that puts the virtual
machine in a known quiesced state, releases its hardware
resources, but keeps the guest and device model in memory so
that forensics can be conducted offline after failover. Polling
HA resource agents should use this new command if postmortem
analysis is important.
We are pursuing a scenario where current polling-based HA resource
agents are complemented with an event-driven failure notification
mechanism that allows for faster failover times by eliminating the
delay introduced by polling and by doing without fencing. This would
benefit traditional software clustering stacks and bring a feature
that is essential for fault tolerance solutions such as Kemari.
Additionally, for those who want or need to stick with a polling
model we would like to provide a virtual machine control that
freezes a virtual machine into a failover-safe state without killing
it, so that postmortem analysis is still possible.
In the following sections we discuss the RAS-HA integration
challenges and the changes that need to be made to each component of
the qemu-KVM stack to realize this vision. While at it we will also
delve into some of the limitations of the current hardware error
subsystems of the Linux kernel.
HARDWARE ERRORS AND HIGH AVAILABILITY
The major open source software stacks for Linux rely on polling
mechanisms to detect both software errors and hardware failures. For
example, ping or an equivalent is widely used to check for network
connectivity interruptions. This is enough to get the job done in
most cases but one is forced to make a trade off between service
disruption time and the burden imposed by the polling resource
On the hardware side of things, the situation can be improved if we
take advantage of CPU and chipset RAS capabilities to trigger
failover in the event of a non-recoverable error or, even better, do
it preventively when hardware informs us things might go awry. The
premise is that RAS features such as hardware failure notification
can be leveraged to minimize or even eliminate service
Generally speaking, hardware errors reported to the operating system
can be classified into two broad categories: corrected errors and
uncorrected errors. The later are not necessarily critical errors
that require a system restart; depending on the hardware and the
software running on the affected system resource such errors may be
recoverable. The picture looks like this (definitions taken from
"Advanced Configuration and Power Interface Specification, Revision
4.0a" and slightly modified to get rid of ACPI jargon):
- Corrected error: Hardware error condition that has been
corrected by the hardware or by the firmware by the time the
kernel is notified about the existence of an error condition.
- Uncorrected error: Hardware error condition that cannot be
corrected by the hardware or by the firmware. Uncorrected errors
are either fatal or non-fatal.
o A fatal hardware error is an uncorrected or uncontained
error condition that is determined to be unrecoverable by
the hardware. When a fatal uncorrected error occurs, the
system is usually restarted to prevent propagation of the
o A non-fatal hardware error is an uncorrected error condition
from which the kernel can attempt recovery by trying to
correct the error. These are also referred to as correctable
or recoverable errors.
Corrected errors are inoffensive in principle, but they may be
harbingers of fatal non-recoverable errors. It is thus reasonable in
some cases to do preventive failover or live migration when a
certain threshold is reached. However this is arguably the job
systems management software, not the HA, so this case will not be
discussed in detail here.
Uncorrected errors are the ones HA software cares about.
When a fatal hardware error occurs the firmware may decide to
restart the hardware. If the fatal error is relayed to the kernel
instead the safest thing to do is to panic to avoid further
damage. Even though it is theoretically possible to send a
notification from the kernel's error or panic handler, this is a
extremely hardware-dependent operation and will not be considered
here. To detect this type of failures one's old reliable
polling-based resource agent is the way to go.
Non-fatal or recoverable errors are the most interesting in the
pack. Detection should ideally be performed in a non-intrusive way
and feed the policy engine with enough information about the error
to make the right call. If the policy engine decides that the error
might compromise service continuity it should notify the HA stack so
that failover can be started immediately.
* Linux kernel
One of the main goals is to notify HA software about hardware errors
as soon as they are detected so that service downtime can be
minimized. For this a hardware error subsystem that follows an
event-driven model is preferable because it allows us to eliminate
the cost associated with polling. A file based API that provides a
sys_poll interface and process signaling both fit the bill (the
latter is pretty limited in its semantics an may not be adequate to
communicate non-memory type errors).
The hardware error subsystem should provide enough information to be
able to map error sources (memory, PCI devices, etc) to processes or
virtual machines, so that errors can be contained. For example, if a
memory failure occurs but only affects user-space addresses being
used by a regular process or a KVM guest there is no need to bring
down the whole machine.
In some cases, when a failure is detected in a hardware resource in
use by one or more virtual machines it might be necessary to put
them in a quiesced state before notifying the associated qemu
Unfortunately there is no generic hardware error layer inside the
kernel, which means that each hardware error subsystem does its own
thing and there is even some overlap between them. See HARDWARE ERRORS IN LINUX below for a brief description of the current mess.
Currently KVM is only notified about memory errors detected by the
MCE subsystem. When running on newer x86 hardware, if MCE detects an
error on user-space it signals the corresponding process with
SIGBUS. Qemu, upon receiving the signal, checks the problematic
address which the kernel stored in siginfo and decides whether to
inject the MCE to the virtual machine.
An obvious limitation is that we would like to be notified about
other types of error too and, as suggested before, a file-based
interface that can be sys_poll'ed might be needed for that.
On a different note, in a HA environment the qemu policy described
above is not adequate; when a notification of a hardware error that
our policy determines to be serious arrives the first thing we want
to do is to put the virtual machine in a quiesced state to avoid
further wreckage. If we injected the error into the guest we would
risk a guest panic that might detectable only by polling or, worse,
being killed by the kernel, which means that postmortem analysis of
the guest is not possible. Once we had the guests in a quiesced
state, where all the buffers have been flushed and the hardware
sources released, we would have two modes of operation that can be
used together and complement each other.
- Proactive: A qmp event describing the error (severity, topology,
etc) is emitted. The HA software would have to register to
receive hardware error events, possibly using the libvirt
bindings. Upon receiving the event the HA software would know
that the guest is in a failover-safe quiesced state so it could
do without fencing and proceed to the failover stage directly.
- Passive: Polling resource agents that need to check the state of
the guest generally use libvirt or a wrapper such as virsh. When
the state is SHUTOFF or CRASHED the resource agent proceeds to
the facing stage, which might be expensive and usually involves
killing the qemu process. We propose adding a new state that
indicates the failover-safe state described before. In this
state the HA software would not need to use fencing techniques
and since the qemu process is not killed postmortem analysis of
the virtual machine is still possible.
HARDWARE ERRORS IN LINUX
In modern x86 machines there is a plethora of error sources:
- Processor machines check exception.
- Chipset error message signals.
- APEI (ACPI4).
- PCIe AER.
- Non-platform devices (SCSI errors, ATA errors, etc).
Detection of processor, memory, PCI express, and platform errors in
the Linux kernel is currently provided by the MCE, the EDAC, and the
PCIe AER subsystems, which covers the first 5 items in the list
above. There is some overlap between them with regard to the errors
they can detect and the hardware they poke into, but they are
essentially independent systems with completely different
architectures. To make things worse, there is no standard mechanism
to notify about non-platform devices beyond the venerable printk().
Regarding the user space notification mechanism, things do not get
any better. Each error notification subsystem does its own thing:
- MCE: Communicates with user space through the /dev/mcelog
special device and
/sys/devices/system/machinecheck/machinecheckN/. mcelog is
usually the tool that hooks into /dev/mcelog (this device can be
polled) to collect and decode the machine check errors.
/sys/devices/system/machinecheck/machinecheckN/trigger can be
used to set a program to be run when a machine check event is
detected. Additionally, when an machine check error that affects
only user space processes they are signaled SIGBUS.
The MCE subsystem used to deal only with CPU errors, but it was
extended to handle memory errors too and there is also initial
support for ACPI4's APEI. The current MCE APEI implementation
reaps memory errors notified through SCI, but support for other
errors (platform, PCIe) and transports covered in the
specification is in the works.
- EDAC: Exports memory errors, ECC errors from non-memory devices
(L1, L2 and L3 caches, DMA engines, etc), and PCI bus parity and
SERR errors through /sys/devices/system/edac/*.
- NMI: Uses printk() to write to the system log. When EDAC is
enabled the NMI handler can also instruct EDAC to check for
potential ECC errors.
- PCIe AER subsystem: Notifies PCI-core and AER-capable drivers
about errors in the PCI bus and uses printk() to write to the
I would appreciate your comments and advice on any of the issues
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to firstname.lastname@example.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/