It didn't make any sense to me for mem(prov) to be the default selector,
since this has too many caveats versus mem(free). Switch to using
mem(free) as the default (i.e. "mem") and make memprov the alternative.
1. Move the test_matrix, volume name, and size to module-level variables
so they can be accessed externally if this is imported.
2. Separate the volume creation and volume cleanup into functions.
3. Separate the individual benchmark runs into a function.
This should enable easier calling of the various subcomponents
externally, e.g. for external benchmark scripts.
1. Error out when trying to add a new network to a VM if the network
doesn't exist on the cluster.
2. When showing the VM list, only show invalid networks in red, not the
whole list.
This replicates some of the more important functionality of the defunct
pvc-flush.service unit. On presence of a trigger file (i.e.
/etc/pvc/autoready), it will trigger a "node ready" on boot. It does
nothing on shutdown as this must be handled by other mechanisms, though
a similar autoflush could be added as well.
Provisions Rocky Linux 8 and 9 systems, and potentially older
CentOS/Fedora/Scientific Linux/SuSE systems. Depends on a custom build
of rinse (3.7.1) with Rocky 9 support.
1. Ensure that system_template and script are not nullable in the DB.
2. Ensure that the CLI and API enforce the above and clean up CLI
arguments for profile add.
3. Ensure that, before uploading OVAs, a 'default_ova' provisioning
script is present.
4. Use the 'default_ova' script for new OVA uploads.
5. Ensure that OVA details are properly added to the vm_data dict in the
provisioner vmbuilder.
This is a breakage between the older version of Celery (Deb10) and
newer. The hard removal broke Deb10 instances.
So try that first, and on failure, assume newer Celery format.
Otherwise the node entries could come back in an arbitrary order; since
this is an ordered list of dictionaries that might not be expected by
the API consumers, so ensure it's always sorted.
This service caused more headaches than it was worth, so remove it.
The original goal was to cleanly flush nodes on shutdown and unflush
them on startup, but this is tightly controlled by Ansible playbooks at
this point, and this is something best left to the Administrator and
their particular situation anyways.
1. Add documentation on the node selector flags. In the API, reference
the daemon configuration manual which now includes details in this
section; in the CLI, provide the help in "pvc vm define" in detail and
then reference that command's help in the other commands that use this
field.
2. Ensure the naming is consistent in the CLI, using the flag name
"--node-selector" everywhere (was "--selector" for "pvc vm" commands and
"--node-selector" for "pvc provisioner" commands).
Adds commands to both replace an OSD disk, and refresh (reimport) an
existing OSD disk on a new node. This handles the cases where an OSD
disk should be replaced (either due to upgrades or failures) or where a
node is rebuilt in-place and an existing OSD must be re-imported to it.
This should avoid the need to do a full remove/add sequence for either
case.
Also cleans up some aspects of OSD removal that are identical between
methods (e.g. using safe-to-destroy and sleeping after stopping) and
fixes a bug if an OSD does not truly exist when the daemon starts up.
With the OSD LVM information stored in Zookeeper, we can use this to
determine the actual block device to zap rather than relying on runtime
determination and guestimation.
Ensures that information like the FSIDs and the OSD LVM volume are
stored in Zookeeper at creation time and updated at daemon start time
(to ensure the data is populated at least once, or if the /dev/sdX
path changes).
This will allow safer operation of OSD removals and the potential
implementation of re-activation after node replacements.
If there is...
1. No '--cluster' passed, and
2. No 'local' cluster, and
3. There is exactly one cluster configured
...then use that cluster by default in the CLI.
Allows an administrator to adjust the PG count of a given pool. This can
be used to increase the PGs (for example after adding more OSDs) or
decrease it (to remove OSDs, reduce CPU load, etc.).
Allows specifying a particular device class ("tier") for a given pool,
for instance SSD-only or NVMe-only. This is implemented with Crush
rules on the Ceph side, and via an additional new key in the pool
Zookeeper schema which is defaulted to "default".
Allows specifying blockdevs in the OSD and OSD-DB addition commands as
detect strings rather than actual block device paths. This provides
greater flexibility for automation with pvcbootstrapd (which originates
the concept of detect strings) and in general usage as well.
This ensures that any client command is logged by the local system.
Helps ensure Accounting for users of the CLI. Currently logs the full
command executed along with the $USER environment variable contents.
Solves two problems:
1. How match fuzziness was used was very inconsistent; make them all the
same, i.e. "if is_fuzzy and limit, apply .* to both sides".
2. Use re.fullmatch instead of re.match to ensure exact matching of the
regex to the value. Without fuzziness, this would sometimes cause
inconsistent behavior, for instance if a limit was non-fuzzy "vm",
expecting to match the actual "vm", but also matching "vm1" too.
This is recommended by the Python Requests documentation:
> It’s a good practice to set connect timeouts to slightly larger than a
multiple of 3, which is the default TCP packet retransmission window.
This allows us to keep a very low connect timeout of 3 seconds, but also
ensure that long commands (e.g. --wait or VM disable) can take as long
as the API requires to complete.
Avoids having to explicitly set very long single-instance timeouts for
other functions which would block forever on an unreachable API.
The Ansible manual can't keep up with the other repo, so it should live
there instead (eventually, after significant rewrites).
The Testing page is obsoleted by the "test-cluster" script.
Allows preserving colour within e.g. watch, where Click would normally
determine that it is "not a terminal". This is done via the wrapper echo
which filters via the local config.
Remove the prepare script, and run the two stages manually. Better
handle Black reformatting by doing a check (for the errcode) then
reformat and abort commit to review.
Instead of requiring the VM to already be stopped, instead allow disable
state changes to perform a shutdown first. Also add a force option which
will do a hard stop instead of a shutdown.
References #148
Make sure all of these move to the root of the repository first, then
return to where they were afterwards, using pushd/popd. This allows them
to be executed from anywhere in the repo.
Instead of erroring, just use the implication that restarting a VM does
not want a live modification, and proceed from there. Update the help
text to match.
Use a power off (and then make the power on a requirement) during a node
fence. Removes some potential ambiguity in the power state, since we
will know for certain if it is off.
Update the docs with the current information on setting up a cluster,
including simplifying the Ansible configuration to use the new
create-local-repo.sh script, and simplifying some other sections.
Refactors some of the code in VXNetworkInterface to handle MTUs in a
more streamlined fashion. Also fixes a bug whereby bridge client
networks were being explicitly given the cluster dev MTU which might not
be correct. Now adds support for this option explicitly in the configs,
and defaults to 1500 for safety (the standard Ethernet MTU).
Addresses #144
When using the "state", "node", or "tag" arguments to a VM list, add
support for a "negate" flag to look for all VMs *not in* the state,
node, or tag state.
1. Remove ramp_time as this was giving very strange results.
2. Up the runtime to 75 seconds to compensate.
3. Print the fio command to the console to validate.
1. Runs `fio` with the `--format=json` option and removes all terse
format parsing from the results.
2. Adds a 15-second ramp time to minimize wonky ramp-up results.
3. Sets group_reporting, which isn't necessary with only a single job,
but is here for consistency.
Allows choosing different list and info functions based on the benchmark
version found. Currently only implements "legacy" version 0 with more to
be added.
Adds a test_format database column and a value in the API return for the
test format version, starting at 0 for the existing format as of 0.9.37.
References #143
1. Move to a time-based (60s) benchmark to avoid these taking an absurd
amount of time to show the same information.
2. Eliminate the 256k random benchmarks, since they don't really add
anything.
3. Add in a 4k single-queue benchmark as this might provide valuable
insight into latency.
4. Adjust the output to reflect the above changes.
While this does change the benchmarking, this should not invalidate any
existing benchmarks since most of the test suit is unchanged (especially
the most important 4M sequential and 4K random tests). It simply removes
an unused entry and adds a more helpful one. The time-based change
should not significantly affect the results either, just reduces the
total runtime for long-tests and increase the runtime for quick tests to
provide a better picture.
Ensure that all keepalive timeouts are set (prevent the queue.get()
actions from blocking forever) and set the thread timeouts to line up as
well. Everything here is thus limited to keepalive_interval seconds
(default 5s) to keep it uniform.
Remove two superfluous synchronization steps which are not needed here,
since the exclusive lock handles that situation anyways.
Still does not fix the weird flush->unflush lock timeout bug, but is
better worked-around now due to the cancelling of the other wait freeing
this up and continuing.
Make the block on stage C only wait for 900 seconds (15 minutes) to
prevent indefinite blocking.
The issue comes if a VM is being received, and the current unflush is
cancelled for a flush. When this happens, this lock acquisition seems to
block for no obvious reason, and no other changes seem to affect it.
This is certainly some sort of locking bug within Kazoo but I can't
diagnose it as-is. Leave a TODO to look into this again in the future.
Rather than using a cumbersome and overly complex ping-pong of read and
write locks, instead move to a much simpler process using exclusive
locks.
Describing the process in ASCII or narrative is cumbersome, but the
process ping-pongs via a set of exclusive locks and wait timers, so that
the two sides are able to synchronize via blocking the exclusive lock.
The end result is a much more streamlined migration (takes about half
the time all things considered) which should be less error-prone.
1. Output from ipmitool was not being stripped, and stray newlines were
throwing off the comparisons. Fixes this.
2. Several stages were lacking meaningful messages. Adds these in so the
output is more clear about what is going on.
3. Reduce the sleep time after a fence to just 1x the
keepalive_interval, rather than 2x, because this seemed like excessively
long even for slow IPMI interfaces, especially since we're checking the
power state now anyways.
4. Set the node daemon state to an explicit 'fenced' state after a
successful fence to indicate to users that the node was indeed fenced
successfully and not still 'dead'.
The previous implementation did not work with /dev/nvme devices or any
/dev/disk/by-* devices due to some logical failures in the partition
naming scheme, so fix these, and be explicit about what is supported in
the PVC CLI command output.
The 'echo | gdisk' implementation of partition creation also did not
work due to limitations of subprocess.run; instead, use sgdisk which
allows these commands to be written out explicitly and is included in
the same package as gdisk.
The default of 0.05 (5%) is likely ideal in the initial implementation,
but allow this to be set explicitly for maximum flexibility in
space-constrained or performance-critical use-cases.
Adds in three parts:
1. Create an API endpoint to create OSD DB volume groups on a device.
Passed through to the node via the same command pipeline as
creating/removing OSDs, and creates a volume group with a fixed name
(osd-db).
2. Adds API support for specifying whether or not to use this DB volume
group when creating a new OSD via the "ext_db" flag. Naming and sizing
is fixed for simplicity and based on Ceph recommendations (5% of OSD
size). The Zookeeper schema tracks the block device to use during
removal.
3. Adds CLI support for the new and modified API endpoints, as well as
displaying the block device and DB block device in the OSD list.
While I debated supporting adding a DB device to an existing OSD, in
practice this ended up being a very complex operation involving stopping
the OSD and setting some options, so this is not supported; this can be
specified during OSD creation only.
Closes#142
Ensures that a VM won't:
(a) Have provisioned more RAM than there is available on a given node.
Due to memory overprovisioning, this is simply a "is the VM memory count
more than the node count", and doesn't factor in free or used memory on
a node, total cluster usage, etc. So if a node has 64GB total RAM, the
VM limit is 64GB. It is up to an administrator to ensure sanity *below*
that value.
(b) Have provisioned more vCPUs than there are CPU cores on the node,
minus 2 to account for hypervisor/storage processes. Will ensure there
is no severe CPU contention caused by a single VM having more vCPUs than
there are actual execution threads available.
Closes#139
Adds a new API endpoint to support hot attach/detach of devices, and the
corresponding client-side logic to use this endpoint when doing VM
network/storage add/remove actions.
The live attach is now the default behaviour for these types of
additions and removals, and can be disabled if needed.
Closes#141
This branch commit refactors the pvcnoded component to better adhere to
good programming practices. The previous Daemon.py was a massive file
which contained almost 2000 lines of direct, root-level code which was
directly imported. Not only was this poor practice, but this resulted
in a nigh-unmaintainable file which was hard even for me to understand.
This refactoring splits a large section of the code from Daemon.py into
separate small modules and functions in the `util/` directory. This will
hopefully make most of the functionality easy to find and modify without
having to dig through a single large file.
Further the existing subcomponents have been moved to the `objects/`
directory which clearly separates them.
Finally, the Daemon.py code has mostly been moved into a function,
`entrypoint()`, which is then called from the `pvcnoded.py` stub.
An additional item is that most format strings have been replaced by
f-strings to make use of the Python 3.6 features in Daemon.py and the
utility files.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
