Update feature/vm-anti-affinity by merging from master #5658
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
fail is always used in combination with Printf.sprintf so combine the two. Signed-off-by: Frediano Ziglio <frediano.ziglio@cloud.com>
More compact code. Signed-off-by: Frediano Ziglio <frediano.ziglio@cloud.com>
"exe" variable is compute in mutiple functions, not very expensive to compute once. Reduce code. Signed-off-by: Frediano Ziglio <frediano.ziglio@cloud.com>
Make sure we release the "sock" file descriptors in all cases. Add test trying to reproduce the issue passing an invalid file descriptor; not a perfect reproduction but failure in this function can happen for instance if daemon is restarted. Signed-off-by: Frediano Ziglio <frediano.ziglio@cloud.com>
We just want information about the current process. Signed-off-by: Frediano Ziglio <frediano.ziglio@cloud.com>
Instrument: - `forkhelpers.ml`, - `fecomms.ml` to create spans around functions when a parent span is supplied as `tracing`. Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
Currently the `observe` mode of the `tracing` library do not work corrently resulting in the logs being spammed by this warning. Comment it out so that the logs do not become too big (for the time being). Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
…/CP-48195-instrument-forkexecd-client CP-48195: Instrument client side of `forkexecd`
- Use `Types.checkError` instead of throwing a generic `XenAPIException`. This ensures the `Types.XYZ` family of exceptions are being used - Use `@JsonValue` to ensure base class objects are deserialised as a simple opaque_ref string, as opposed to a mapping of each field. This ensures the API's behaviour is unchanged. - Parse the results of `task.getResult` calls. The jsonrpc method returns value payloads of the form `"value" : "<value>OpaqueRef:XYZ</value>"` so we need to strip the surrounding XML Signed-off-by: Danilo Del Busso <danilo.delbusso@cloud.com>
…e build @check Bytecode builds for `http_lib` are disabled due to '(modes best)', and that means that anything that depends on it must have it disabled too to avoid this warning. Avoids these kinds of warnings: ``` File "_none_", line 1: Error: Module `Buf_io' is unavailable (required by `Http_svr') ``` Signed-off-by: Edwin Török <edwin.torok@cloud.com> Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
…ries There were 3 modules with conflicting names with compiler libraries: Watch, Debuginfo and Stats. Debuginfo was renamed, the others's libraries were changed to be wrapped. Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Pinning the libraries runs dune subst, which needs a project name, define it Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
It is not safe to access a global hashtable from multiple threads, even if the operations are read-only (it may be concurrently changed by another thread, which then may result in errors in the racing thread). This means we must always take the mutex, and because OCaml doesn't have a reader-writer mutex, we need to take the exclusive mutex. Eventually we should use a better datastructure here (immutable maps, or lock-free datastructures), but for now fix the datastructure that we currently use to be thread-safe. Signed-off-by: Edwin Török <edwin.torok@cloud.com>
In preparation for OCaml 5, on OCaml 4 they'd be equivalent. Note that adding Atomic doesn't make operations on these values always atomic: that is the responsibility of surrounding code. E.g. Atomic.get + Atomic.set is not atomic, because another domain might've raced and changed the value inbetween (so in that case Atomic.compare_and_set should be used). However for global flags that are read multiple times, but set from a central place this isn't a problem. Signed-off-by: Edwin Török <edwin.torok@cloud.com>
Sets `observe` to `false` until at least one tracer provider is enabled. If all tracer providers are disabled we set it back to `false`. Prior to this `observe` seemed to be always `true` (apart from unit tests). This would cause `with_tracing` to spam the logs with warnings `No provider found...` until at least one tracer provider is enabled. By setting `observe` to `false` as default and updating it depending on the state of the tracer providers, `with_tracing` should now execute no extra operations. Therefore, we avoid spamming the logs unnecessarily. Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
Do not output loglines that are part of the normal operation. Use debug for them, they are not usually logged, but can be enabled if need be by changing the loglevel Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
New version got released Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
The query on HTTP endpoint /updates will return the available updates in JSON format. Prior to the changes in this commit, if a query arrives when another query is being handled, a "GET_UPDATES_IN_PROGRESS" error will be returned immediately. This behaviour is not friendly to GUI client XenCenter. In this commit, the behaviour is changed to wait and retry in handling the query in xapi since the "*_IN_PROGGRESS" error is a transient failure. Tolerating it in xapi (server) side avoids error handling in client side. With the change, the "GET_UPDATES_IN_PROGRESS" will not be an error exposed to users any more. Therefore it is removed. Signed-off-by: Ming Lu <ming.lu@cloud.com>
…389319 CA-389319: Wait and retry for GET_UPDATES_IN_PROGRESS
As part of a start, resources are allocated for a VM in "scheduled_to.." fields. These need to be cleared if the start fails. It turned out that this was incomplete for PCI slots and those were leaking. This patch tries to be more systematical about it. Signed-off-by: Christian Lindig <christian.lindig@cloud.com>
[ASAP] CA-392453: Misc fixes to Java SDK
Create a new unit test file `test_tracing.ml` for the `tracing` library. Add tests for `create`/`set`/`destroy` tracer providers. Now that we change the library to have `observe` modes based on whether or not we have `tracer providers` enabled, we want to make sure that functions on applied on tracer providers set the correct mode for the library. Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
Removes code duplication in `storage_mux.ml` by using the already existing `with_dbg` implementation from `debuginfo` module. This should lower the chances of unintentionally introducing bugs by having to maintain two version of the same functions. e.g. Not using the no op when tracing is disabled and generating unwanted warning messages. Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
Moves the following: - `create` under `TracerProvider`; - `set` under `TracerProvider`; - `destroy` under `TracerProvider; - `get_tracer_providers` unde `TracerProvider`; - `get_tracer` under `Tracer`. Adds documentation for `TracerProvider` module. It kept being confusing of what `Tracing.set` does unless I was going through the implementation again and again. Therefore, I moved some of the functions so that their functionality becomes (hopefully) more intuitive. Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
This is mostly copied from the jekyll version at xapi-project.github.io and amended for Hugo. The main difference is that all XenAPI pages have now been integrated with the main menu in the left side bar, including the menu of the XenAPI classes. Some automation is left to do in order to take the `doc/data` files from a build of `ocaml/idl` and update the class pages using the script `doc/make-class-pages.py`. Also the overview of xapi releases is still missing, as well as some of the API guides. Finally, there is some overlap with markdown files in `ocaml/doc`, which needs to be sorted out. Signed-off-by: Rob Hoes <rob.hoes@citrix.com>
From Xen 4.19 onwards the legacy paths disappeared and the only valid path for pygrub is /usr/libexec/xen/bin/pygrub. This path has always been preferred, but now it's mandatory. Signed-off-by: Alejandro Vallejo <alejandro.vallejo@cloud.com>
Add a new field `cluster_stack_version` to the cluster datamodel to track the version of corosync currently in use. This version will always be set to 3. Also add logic to switch corosync binary and associated library versions when a cluster is created, if needed. Signed-off-by: Vincent Liu <shuntian.liu2@cloud.com>
In case of getting the list of files we called Unix.readlink twice for every file descriptor. In case we wanted just the count we computed the list anyway. Factor out code to avoid calling function twice and do not compute the list is not needed. Signed-off-by: Frediano Ziglio <frediano.ziglio@cloud.com>
The VM power state should be preserved on a suspended VM import. In commit ebb58a8, the power state and suspend VDI in `vm_record` were reset on importing a suspended VM. This meant to facilitate the following `Client.VM.create_from_record`. But this gets the restore of power state and suspend VDI broken as it requires the data in `vm_record`. This commit preserves the data in `vm_record` and reset it just before it being passed to `Client.VM.create_from_record`. Signed-off-by: Ming Lu <ming.lu@cloud.com>
CA-392930: Fixed exception handling which prevents connection
Instruments `xapi_session.ml` to create span around session logins. Adds new attribute for spans: - `xs.xapi.session.originator` where available. Signed-off-by: Gabriel Buica <danutgabriel.buica@cloud.com>
…orosync3-basic CP-48027: Corosync upgrade add `cluster_stack_version` datamodel change
…392836 CA-392836,CA-392847: Lost the power state on suspended VM import
This forces the failure on a host that is trying to perform corosync upgrade. There are ways to recover: if the failure happens early, before the cluster is created in the DB, then a recreate ought to fix the problem. This happens when the corosync upgrade fails on the coordinator. If the failure happens after the cluster is created on a pool member, then a `pool-resync` should help retry this upgrade. Hopefully this can simulate some of the failure paths, but is by no means exhaustive. Other more complicated failures are not easily recoverable and therefore not simulated for now. Signed-off-by: Vincent Liu <shuntian.liu2@cloud.com>
Refactor the representation of Kerberos Domain Controller (KDC) to better support IPv6. It became apparent that the current implementation assumes standard port 88 being used by a KDC - which we should relax in the future. Capture a KDC configuration as KDC.t and provide functions to create, serialise, log a value. Represent the IP address as Ipaddr.t (which can be IPv6 or IPv4). The existing implementation does not properly use a domain controller's port number that it obtains from a `net lookup kdc`: * the port number is not passed to other net commands that query the domain controller * the port number is not stored in the xapi database (along the IP address). Signed-off-by: Christian Lindig <christian.lindig@cloud.com>
We want to store the KDC address as IP/port in xapi using the existing database field. We use a new custom URI. We support the existing (simpler legacy) scheme by trying it first when readign the field. Signed-off-by: Christian Lindig <christian.lindig@cloud.com>
…/CP-48195-tracing-observe-modes
Remove possible file descriptor leak if safe_close_and_exec fails
…orosync3-upgrade-fist CP-49635: Add FIST point for corosync upgrade
This unblocks the Storage: BST to use distributed tracing as the sr_uuid parameter is sometimes not given. Signed-off-by: Steven Woods <steven.woods@citrix.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
It's unused outside of the repository Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
Signed-off-by: Pau Ruiz Safont <pau.ruizsafont@cloud.com>
…-391381 CA-391381: Avoid errors for Partial Callables in observer.py
Remove CVM and relevant test cases CVM had been supported but not since long time ago. This commit is to clean up the remanent code and test cases. Signed-off-by: Ming Lu <ming.lu@cloud.com>
|
minglumlu
approved these changes
May 30, 2024
robhoes
approved these changes
May 30, 2024
robhoes
merged commit 3dc2b9e
into
xapi-project:feature/vm-anti-affinity
14 checks passed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.