IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
It's possible to run the `vmcheck` tests against an existing VM which
may not have any existing `/etc/yum.repos.d` directory. Since the
tests are providing their own repo files, we should be able to run
them in this uncommon configuration.
Closes: #1530
Approved by: jlebon
Prep for rojig-only compose support. We want to be able
to quickly detect whether or not there were changes in the
input, just like we do when targeting an OSTree repository.
Closes: #1528
Approved by: jlebon
I'm trying to have a more opinionated model where custom builds
use inheritance, and currently one can only have a single
`postprocess-script`.
Further, in YAML it's very convenient to use inline vs external
data.
Closes: #1527
Approved by: jlebon
Similar to the passwd conversion, this commit
implements conversion of group entries into sysuser struct entries
Tests are also included in the commit
Closes: #1519
Approved by: cgwalters
It turns out we basically have to slap an `Option<T>` around
everything, (in particular `bool` etc.) we need to be able
to distinguish in (I believe) all the cases between
"value unspecified" and "value provided".
Concretely it didn't work to try to set `machineid-compat: false`
in an included yaml treefile becuase it was just defaulted to `true`
by the toplevel.
Down the line we should move all of the parsing into Rust
and have two different `struct` types for "YAML we load" versus
"verified treefile".
Closes: https://github.com/projectatomic/rpm-ostree/issues/1524Closes: #1525
Approved by: lucab
This teaches the client to fetch packages from URLs directly so that one
doesn't have to `curl` first and then install. Supported anywhere
package filenames are allowed (notably: `install` and
`override replace`).
One neat things about this is that we download the file into an
`O_TMPFILE` and then pass on ownership of that fd directly to the
daemon. So at no point are the packages actually laying visible on the
system. (Assuming the filesystem supports `O_TMPFILE` that is).
This adds direct linking to libcurl and openssl, two libraries which we
were already pulling in indirectly.
Closes: #1508
Approved by: cgwalters
This is for: https://github.com/projectatomic/rpm-ostree/issues/718
But I'm not going to close that issue as this only does the server
side, and I think we should support it client side too.
Since I wrote that issue, we ended up skipping the `dnf_transaction_depsolve()`
API, and hence we don't need to block on a libdnf change. So
this was quite simple.
Closes: #1513
Approved by: jlebon
We've put a lot of work into staged deployments, it's time
to pull the trigger and turn them on by default. This is
a key step for enabling `stage` mode automatic updates by
default in e.g. Fedora CoreOS/Silverblue.
We add a new `--disable-staged` build-time option to flip
things back.
Closes: #1430
Approved by: jlebon
Staging fixes the `/etc` bug for livefs. There's actually more
we could do here around taking advantage of staging for livefs;
for example, I think once the livefs is complete, we could just delete
the staged deployment. And then we don't need to render on the next
boot the live status, etc.
Anyways, all that can come in the future. This is prep for
enabling staging by default.
Closes: #1430
Approved by: jlebon
Fix `--booted` was printing an extra line after `Deployments:`. While
we're there, also update the section name to `BootedDeployment:` in that
case to be more correct. (We're not printing all the deployments, only
the booted deployment.)
Closes: #1503
Approved by: cgwalters
The `f28-compose` test keeps timing out. Some time recently, I/O
performance of the internal OpenStack instance used for testing has
degraded. I have a ticket open to investigate the regression though
haven't had any luck so far.
Let's just take the easy way out and split the test into two testsuites.
This is obviously hacky, and sad, and unfortunate. But the PRs must keep
flowing until we finally wean off of OpenStack.
Closes: #1498
Approved by: cgwalters
And don't mention "replace" in the error texts.
This is (now) known broken with kernel updates; making that not
dangerous isn't going to be easy. I debated entirely removing it
but for now let's just make it harder to use.
Ref: https://github.com/projectatomic/rpm-ostree/issues/1495Closes: #1497
Approved by: jlebon
We should be expecting testpkg-1.1-1 here, not 1.0-1. This was passing
before because of the nondeterministic `find` output (fixed in the
previous commit) which could spit out the older rojig RPM.
Closes: #1491
Approved by: cgwalters
This test was relying on the order in which `find` reports matching path
names to find the right RPM. This was failing for me locally sometimes
because it matched the wrong RPM file. Fix this by just directly
referencing the full path name since we can.
Closes: #1491
Approved by: cgwalters
My fix to the testsuite in #1488 in which I made the `machineid-compat`
test part of `test-basic.sh` wasn't correct since the basic tests in
`libbasic-test.sh` also check that the default behaviour without the
`machineid-compat` option is to include it.
Let's just do this right and split out the `machineid-compat` test into
its own run.
Closes: #1491
Approved by: cgwalters
I noticed that the latest Fedora Atomic Host 28 and Silverblue did not
have an `OSTREE_VERSION` line in `/etc/os-release` even though both
specified `mutate-os-release` in their manifests. This turned out to be
due to the fact that `/usr/lib/os-release` is now a symlink to a
variant-specific file (e.g. `os-release-atomichost`), so we would
fallback to mutating `/usr/lib/os.release.d/os-release-fedora` instead.
Fix this by just taking the nuclear option of running `realpath` in the
rootfs directly. This is more maintainable than trying to keep up with
changes in variants/naming/etc. There's related discussions to this in
the original [PR](https://github.com/projectatomic/rpm-ostree/pull/410)
which introduced the feature re. resolving symlinks within the rootfs.
Closes: #1481
Approved by: jlebon
Rather than silently ignoring it.
In theory...we could write to /usr/lib/systemd instead of `/etc`
but eh...I feel like what we really want to do is make it convenient
to write a preset file from the YAML.
(We could have an `add-files` content that takes values literally
which would be nice in YAML and suck in JSON)
A general thread running through this is that for people making
*derivatives* of a CoreOS-like system, having to create their
own `exampleos-release` package is an annoying hurdle.
Anyways for now we're fixing the bug that we were silently ignoring
it.
Closes: #1488
Approved by: jlebon
The rojig spec is almost entirely rpm-ostree implementation details;
let's not have lots of people fork/duplicate it. Rather add the bits
of rojig to the treefile that people need to define (most notably
the name).
Prep for stabilizing rojig.
I had a few false starts with this PR; managing ownership/lifetimes
across C/Rust is just complicated. I got bit hard by the fact that
the workdir in `--unified-core` is really dfd-relative, and had to
do a dance to propagate the dfd into rust, as well as down into
the rojig builder.
Closes: #1484
Approved by: jlebon
Make logging work the same as it does for the vmcheck-STI work
(at some point I'll try to unify the 3 parallel+script implementions
we have). This fixes the problem that when the test times out,
the filename won't have `.txt` and S3 won't have the right MIME type.
Closes: #1479
Approved by: jlebon
This is analogous to `upgrade --upgrade-unchanged-exit-77`, but for
`install`/`uninstall`. This way, one can determine whether the command
truly had an effect on the default deployment or not.
Since this works by comparing against the previous default deployment,
this works correctly even if we already had a pending deployment at the
beginning of the transaction.
Closes: #1478
Approved by: cgwalters
Add a new `install/uninstall --idempotent` option to make it easier to
interact with the CLI through scripts. E.g. one doesn't have to check
first if a request has already been installed/uninstalled.
Closes: #1467Closes: #1478
Approved by: cgwalters
This makes it easier for scripts to determine whether there is a pending
deployment instead of using `--json/--jsonpath`.
Closes: #1478
Approved by: cgwalters
The `cached-update` variant would mark a bunch of RPMs as upgraded even
if they didn't actually change. The issue turned out to be we were doing
the diff all wrong in the staged deployment case. I'm not sure what I
was thinking in #1344, but essentially, we were marking all layered RPMs
in the staged deployment as updates instead of only marking those
layered RPMs which were actually changed EVR.
We just simplify the approach here by directly doing a pkglist diff
between the booted and staged deployments and consuming that. That's
really all there is to it! Reduces the code quite a bit too.
Closes: #1446Closes: #1455
Approved by: cgwalters
This one snuck in as part of #1205. I had initially started writing
`inject-pkglist.c` in Python and somehow inadvertedly made it part of
a commit.
Closes: #1455
Approved by: cgwalters
In line with the recent trend of marking things stable, and in
preparation for stabilizing `rojig://` - Let's stabilize the `--unified-core`
option for `compose tree`.
I'm not sure we could make it the default anytime soon; today it trips
over bugs in the PAM package in RHEL7 for example. But it
works fine for Fedora, and I think the code/design are good enough to be stable.
Closes: #1465
Approved by: jlebon
Follow up to: https://github.com/projectatomic/rpm-ostree/pull/1459
We now honor arch-specific packages in YAML, and reject unknown
architectures. I looked a little bit at how to avoid having hardcoded
arch lists, but it doesn't seem worth it right now.
Closes: #1468
Approved by: jlebon
Let's not make the same mistake we did with JSON where typoing a
field means it's silently ignored. This actually caught a bug
in a YAML usage we had:
```
error: Failed to load YAML treefile: unknown field `install_langs`, expected one of ... `install-langs` ...
```
Yes, this is a compatibility break with the feature we just announced
but...I seriously doubt anyone (that isn't known to me) has converted
yet, and if they are excited enough to start using a two-week-old feature
they can adjust.
Closes: #1459
Approved by: cgwalters
We're looking to embed an ostree commit inside a container image,
to make it easier to transport around with other images.
Conceptually here the host system is tracking a container (just
like for rojig we're tracking an RPM). This is the first step
towards making that support nicer; tooling can do
`rebase --custom-origin-url oscontainer://quay.io/exampleos@sha256:...`
and have that show up in `rpm-ostree status`.
There are two values, one intended to be machine readable (like
the `ostree://` and `rojig://` and one for humans which we
display when an admin types `rpm-ostree upgrade`.
This builds on prior work in
27bd7b97bb from #1396 .
Closes: #1406
Approved by: jlebon
This test was hard to adapt for staging. I did the obvious change
of just rebooting after most of the `kargs` invocations, but much
later in the suite there was a test that compared vs the "pristine"
kargs. Supporting that would be a bit hard...I briefly thought
about pinning etc. but it gets so messy to do in both staged and not-staged
defaults. So I just deleted those bits.
Closes: #1453
Approved by: jlebon
Noticed this while looking at the logs for #1432.
Because --check and --preview exit with rc=77 when there are no updates,
we would actually stop early on in the test and marking it as SKIPPED.
Fix this by making sure we explicitly check for the $rc we expected when
using those switches.
I also added a final grep pass to make it easy to inspect whether we
skipped any tests. I was about to do this nicely in `multitest.py`
instead, though it may not be of this world much longer, so meh...
Closes: #1450
Approved by: cgwalters
First, split it into its own section; it's important enough to merit it.
Second, explicitly reference the systemd timer/service units. For
example, a question I often have is "when is the next run" and of course
you can get that rpm `systemctl status rpm-ostreed-automatic.timer` but
you have to know that, and the reminder helps.
(I briefly looked at implementing the `Trigger` line from `systemctl status`
but it's not entirely trivial...tempting to just fork off a `systemctl status | grep `)
Prep for unifying this text with the message we print when one does
`rpm-ostree upgrade` when auto-updates are enabled.
Closes: #1432
Approved by: jlebon
First the pinning tests would try to pin a staged deployment,
and some of the later tests here depend on a subtle way on the
state of the system. It's tempting to do a `reset` before each one
and reboot but this makes things work.
There's some additional assertions here as I went through and
was debugging.
Prep for making staging the default.
Closes: #1438
Approved by: jlebon
Add a new `reset` command that makes it easy to blow away all
customizations: overlays, overrides, and initramfs. One can use flags to
only reset some of the customizations.
I placed this under `ex` out of conservatism. It's a pretty simple
command with simple behaviour, though the features it relies on
(no-layering, no-initramfs) are brand new. We can move it out of there
in a release or two?
Closes: #1387Closes: #1419
Approved by: cgwalters
It's been in experimental for a while now, and we've had good feedback
that it's working. With #1392 fixed, it seems in a good position to
declare stable and commit to that API. This also helps empty out `ex` a
bit.
Closes: #1428
Approved by: cgwalters
Follow-up improvement after
https://github.com/openshift/os/pull/135
This should ensure it survives systemd's preset run on firstboot.
Although honestly...what we should *really* do is check whether
the `default.target` symlink target exists, and if not reset it
to `multi-user.target` so no one would have to care, but that'd
be conceptually separate from this, so I may do it later.
Closes: #1427
Approved by: jlebon
We actually want systemd's `ConditionFirstBoot` to fire. The
primary rationale here is that we're adopting Ignition for Fedora CoreOS,
and having `ConditionFirstBoot=` function will help a lot, as the idea
is it only runs once.
However, I discovered that this breaks the `units` directive for example,
as systemd blows away all the unit state in `/etc`. The correct thing
to do from the start is to use presets. We could add an implementation of
`units` which works with this on and instead writes a preset file but...eh.
My plan is to at some point introduce an "epoch" and flip various defaults,
this one, `tmp-is-dir`, the passwd file handling, etc.
See: https://github.com/dustymabe/bootengine/pull/11Closes: #1425
Approved by: jlebon
We added the `no-layering` option, but made it conflicting with
`--install`. This loosens that requirement so that one can do e.g.
rpm-ostree uninstall --all -install foobar
to essentially remove all layered packages and then add back `foobar`.
Prep for `reset` command.
Closes: #1424
Approved by: cgwalters
Kernel arguments added from `rpm-ostree ex kargs` can be lost
after doing multiple deploy related operations (e.g upgrade, install etc).
The root cause of that is because when doing the deployment
(ostree_sysroot_deploy_tree), the kernel arguments from the
pending deployment were never passed in.
This commit adds logic to pass in kernel arguments from pending deployment
to fix the above mentioned problem
Closes: https://github.com/projectatomic/rpm-ostree/issues/1392Closes: #1420
Approved by: cgwalters
This is one part of #1387 and is prep for a single higher level command
that makes it easier to remove all customizations (layering *and*
overrides).
Closes: #1408
Approved by: cgwalters
Doing `override reset --all` may be a no-op if there are no overrides
currently applied. But we were trying to always print the tree diff
under the assumption that all override invocations have to result in a
new deployment.
Fix this by using the same default deployment check that `upgrade` does
before printing.
Closes: #1407
Approved by: cgwalters
This fixes a painful UX issue where one must use the full NEVRA when
uninstalling a locally layered RPM. Now, one can specify either the
NEVRA or the package name only. Though we still try to interpret the
request first as a NEVRA or a capability.
Closes: #1386Closes: #1400
Approved by: cgwalters
There are a few scenarios today where one might deliver content
to a machine via an external transport. For example, take the
scenario of a single server updated via USB drive. While we
can provide a refspec...what should the remote be? (This gets
into ostree collections). There's nothing really that can
happen when typing `rpm-ostree upgrade` unless the USB stick
is plugged in. That type of scenario should be emphasized
by pinning the commit - the machine is updated via an external
script.
Another case: we're experimenting embedding OSTree commits inside OCI
containers. Here again since rpm-ostree can't understand how to
pull content from containers, it's saner to drop the refspec
bits, and pin to a commit.
Further enhancements will follow to make the admin experience more
obvious.
Closes: #1396
Approved by: jlebon
When I initially added support for local RPM layering (#657), I was
operating under the assumption that there is no reliable way to parse a
NEVRA back to its constituent elements. So we worked around this by
doing lookups in the pkgcache and comparing against branches converted
back to their NEVRAs, which was pretty hacky.
Later on, I learned that you *can* reliably parse a NEVRA back to its
elements because RPM forbids all fields from having a colon `:` and all
fields other than the package name to have a dash `-`. In fact, there is
a helper from `libdnf` to do exactly this.
This patch swaps out the old way of decomposing NEVRAs and converting
NEVRAs to their cache branch notation with a simpler more direct
approach using the helper function from `libdnf`.
Closes: #1395
Approved by: cgwalters
For now all this switch does is turn off the scary warning. We also
tweak the warning to make it clear that this will be required in a
future release.
Closes: #1378
Approved by: cgwalters
Let's modernize and start supporting YAML treefiles. I'll dare make the
sweeping generalization that most people would prefer reading and
writing YAML over JSON.
This takes bits from coreos-assembler[1] that know how to serialize a
YAML file and spit it back out as a JSON and makes it into a shared lib
that we can link against. We could use this eventually for JSON inputs
as well to force a validation check before composing.
If we go this route, we could then turn on `--enable-rust` in FAHC for
now and drop the duplicate code in coreos-assembler.
[1] https://github.com/cgwalters/coreos-assemblerCloses: #1377
Approved by: cgwalters
I was honestly surprised to see this test was running all this time for
CentOS. As cool as it is, this test is really more meant for Fedora.
Closes: #1382
Approved by: jlebon
We were trying to test that doing `override remove` on a layered package
errored out. But the test was bogus since `foo` is a base package here.
Let's just use a brand new package that will actually be layered.
Closes: #1382
Approved by: jlebon
Just part of the continuing breakdown of our testsuites into smaller
pieces for more efficient allocation.
Will add a test case to the `layering-basic-2.sh` in a follow-up commit.
Closes: #1382
Approved by: jlebon
With the new support for pinning deployments, we need to also update
rpm-ostree to clean up the transient state as is now done in the ostree
sysroot upgrader.
This addresses that issue as well as tries to be a little cleaner in how
we clean up other transient state. Notably, we add a new helper function
to `RpmOstreeOrigin` to do this for us and use it in the upgrader. In
other cases, we do want this transient information since it allows us to
describe the deployment.
Closes: https://github.com/ostreedev/ostree/issues/1595Closes: #1372
Approved by: cgwalters
This makes the logs a bit more useful, but the ultimate goal
here is to write the originating client `id` to the cached update
data, so users know that e.g. `gnome-software` triggered it.
Closes: #1368
Approved by: jlebon
Following up to https://github.com/projectatomic/rpm-ostree/pull/1352
AKA 506910d930
which added an experimental flag to globally enable deployment
staging, let's add an `ex-stage` automatic update policy.
I chose to create a new `test-autoupdate-stage.sh` and rename
the previous one to `test-autoupdate-check.sh` in going with
the previous theme of smaller test files; it's
way faster to iterate on new tests when it's a new file. And adding
staging at the top would have been weird.
This was all quite straightforward, just plumbing through lots
of layers.
Closes: #1321
Approved by: jlebon
The common case is having layered packages and no overrides; seeing
`Applying 8 overlays` then `Building filesystem [0/8]` is redundant.
Tweak the progress to avoid the double message. Also change the terminology
to clarify that each item is a package.
Also update libglnx to pull in progress bar newline fix.
Update submodule: libglnx
Closes: #1363
Approved by: jlebon
Follow-up from #1344.
In the case where a cached update is created from an `upgrade` operation
(and soon, "stage" auto-update policy runs), we can just print the diff
and advisory info together with the pending deployment. This makes the
output look much more natural.
Closes: #1350
Approved by: cgwalters
This now works after https://github.com/ostreedev/ostree/pull/1570
However, I was hitting an issue with the temp httpd running the
tests serially, let's split them up.
Really the only sane thing is to run each of these tests in a fresh
VM; we'll do the VM-in-container pattern here hopefully soon.
Closes: #1355
Approved by: jlebon
Now that infrastructure for this has landed in libostree,
let's make it easy for people to opt-in to testing it. This is a distinct first
step for adding it as an update policy.
Closes: #1352
Approved by: jlebon
Right now, cached updates generated during "check" policy runs are
completely decoupled from upgrade operations. This can lead
to the surprising situation where the "Available update" is *older* than
a freshly deployed pending tree with `rpm-ostree upgrade`.
We should just generate a cached update after upgrade operations. This
is also prep for staged deployments, where we'll want to do this as
well.
Note that we write out the cached update here even if automatic updates
are turned off since it's essentially free. I've been thinking about
always displaying that information after an `rpm-ostree upgrade` in
`status`. Though not sure if we should keep it in a separate "Available
update" section, or somehow morph it as part of the pending deployment
output.
Closes: #1344
Approved by: cgwalters
Previously we merged: #1228 AKA 12dc565b00
My recollection is that was working on it the background, while doing
something else, and I clearly didn't get to the point of testing it "for real".
There are many interlocking issues here to make this work. For example,
the "remove RPM" logic needs special handling for the kernel, because
we also inject content into `/usr/lib/ostree-boot` and also generate
the initramfs, etc.
The architecture I chose is to have the core *detect* when a kernel
is changed, and also call into the kernel processing code when removing
a kernel package. But the logic for doing kernel reinstallation client-side
is best alongside the initramfs generation logic which already existed
in the sysroot upgrader.
I extended the test suite to cover what was failing before, and I
tested this interactively. But I'm uncertain about adding a test
for actually *booting* into the GA kernel as it's quite possible
some bits in userspace rely on a newer kernel. Fixing this properly
really wants some infrastructure to better "re-version" an existing
package without changing its content.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1334Closes: #1346
Approved by: jlebon
Now we stop running rpm-ostreed as non-root, which is going to be
a maintenance pain going forward. If we do introduce non-VM based
tests I think we should look to doing in-container testing.
Closes: #1339
Approved by: cgwalters
Building on:
- 9cbec27d4c
- e7a42f70a9
I was looking at a rpm-ostree run that imports a variety of rpmmd-repos,
and information about the source repositories is really useful for determining
the up-to-dateness. We've been capturing this data for a while, it's
about time we started showing it somewhere.
This does make `status --verbose` notably more verbose, but eh, that
seems fine for now.
See also https://github.com/projectatomic/rpm-ostree/issues/774Closes: #1345
Approved by: jlebon
It's very normal for base packages to own directories with
dependent packages installing files there. Doing an rm-rf
for directories was just wrong. Concretely this fixes
an `override replace ./systemd-*.rpm`.
librpm is also pretty conservative here (for good reason)
and just ignores `ENOTEMPTY`, so let's match that.
I opted to split things up so we remove not-directories
in a first pass, then remove all directories we can in the
second. This should maximize our chances of removing what we can
in a scenario where e.g. two co-dependent packages install files
to a directory one of them owns.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1340Closes: #1341
Approved by: jlebon
Following up to: https://github.com/projectatomic/rpm-ostree/pull/1336
It makes sense to keep the library tests as unit tests (although
we should also support doing them installed).
The upgrade-rebase tests will move into vmcheck/ soon.
Closes: #1338
Approved by: jlebon
Our test suite originated when package layering was still being
developed, but now that that's mature, the logic where layering
tests are distinct makes less sense.
The `basic` test had grown to really be a collection of many
miscellaneous things. Let's make that more explicit. Further,
let's avoid having each test suite grow too large; when a single
test fails we don't have an easy way to rerun just that test,
so a crude way to have faster local iteration is to split into groups.
My plan is to reintroduce a `basic` test that covers the basics
of all functionality - update, deploy, layering, etc. The advanced/corner
cases of layering like the `rm -rf /` test would still live in a
`test-layering.sh` or so.
Closes: #1336
Approved by: jlebon
The unit tests run an rpm-ostree daemon as non-root, which worked
surprisingly well for quite a while. But it started failing when
working on a patch which adds caching that writes to `/var`.
Since we have the vmcheck system now, let's switch over to that.
This PR moves the random "basic" tests we'd accumulated like
one to verify `StateRoot` is only in `status --verbose`, but not
the tests for the `rebase` command etc.
Closes: #1336
Approved by: jlebon
We should be checking that there are no requested base replacements
here, not removals. This was likely a copy-paste slip-up from copying
the override-remove tests. I still kept those checks for sanity, and
just tacked on the correct ones.
Closes: #1323
Approved by: cgwalters
This fixes a small regression from #852 which prevented inactive
overrides to be reset. Which is funny, because that's exactly the most
likely time when you would want to reset an override.
Basically, the `!is_layered --> no overrides` doesn't make sense for
inactive overrides. I suspect most people worked around this by just
using `reset --all`.
Closes: #1323
Approved by: cgwalters
When juggling multiple test VMs for different purposes, it's useful to
be able to easily e.g. `make vmcheck` a specific one by overriding the
ssh-config file to use, rather than editing the latter each time.
Closes: #1324
Approved by: cgwalters
I was trying a `--ex-unified-core` compose of FAW, and things fell over
on `urw-base35-fonts` which does a dance of setting a stamp file in
`%post` and checking it in `%posttrans`.
This whole pattern should be considered deprecated by file triggers. But let's
support it for now.
Note there's a lot of parameter passing as we need a single directory which is
held across multiple script invocations.
Closes: #1319
Approved by: jlebon
Specifically in this case, this allows us to close a race condition
during `upgrade --check` where the `CachedUpdate` property might not
have been updated yet when we read it after finishing the transaction.
Closes: #1311
Approved by: cgwalters
In this PR: https://github.com/projectatomic/rpm-ostree/pull/1309
I was hitting race conditions running `ostree admin pin` then
`rpm-ostree cleanup` as it was possible that the daemon hadn't handled
the inotify on the sysroot and reloaded the deployment state before
the txn request came in.
Close this race by doing an implicit `reload` before starting a txn.
This is a pretty efficient operation because for the sysroot we're
just doing a `stat()` and comparing mtime.
Implementation wise, change the external API to drop the "did change"
boolean as nothing outside of the `sysroot.c` file used it.
A followup to this would be changing the `status` CLI to call a
(new) DBus API like `RequestReload` that at least did the sysroot
reload if the daemon was otherwise idle or so? And it'd be available
to unprivileged users.
Closes: #1311
Approved by: cgwalters
Moving to caching the GVariant to disk rather than during RPMOSTreeOS
reloads broke the legacy `DownloadUpdateRpmDiff` path because it relied
on the implied recalculations that occur on reloads to update
`CachedUpdate`.
This patch series was initially going to be about just migrating all the
legacy APIs to make use of the new metadata, which would have fixed this
properly. But we first need some real coverage in that aread, which is
very poor right now. I'd like to investigate integrating the Cockpit
tests (at least the ostree-specific parts) into our CI to remedy this.
Anyways, for now at least, let's fix Cockpit.
Closes: #1300Closes: #1303
Approved by: cgwalters
This patch teaches `db list` to also use the pkglist metadata when it
can, just like we did for `db diff`. To spell it out: this then allows
`db list` to work on commits for which we only have the commit object.
I went for the surgical incision here and didn't try to support
invocations which use fnmatch patterns for now. Definitely possible,
though it didn't feel like it was worth the effort given that the common
case is just a raw `db list` (I'd wager most people are probably
hard-wired to pipe to `grep` anyway for filtering).
Also fix the usage string, which had the arguments flipped.
Closes: #1299
Approved by: cgwalters
Support a `--base` switch to perform the diff on the base layer if
deployment checksum was locally composed. This is useful to filter out
from the diff changes due to layered packages.
Closes: #1299
Approved by: cgwalters
Because otherwise, there's no way to see the output of the script.
Also, turn off `gather_facts` since in the majority of cases, we don't
need it, so let's avoid the overhead. We can make this an opt-in flag
later on if needed.
Closes: #1304
Approved by: cgwalters
This came up in a few places recently; it happens for RHEL in some
cases, and in general we don't want to completely fail the daemon
start if someone messes up their remote config.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1301Closes: #1302
Approved by: jlebon
This adds a shell primitive to make it easy to execute a playbook
task list.
The big picture idea is to sync with https://github.com/ostreedev/ostree/pull/1462
and rewrite some of the libvm shell stuff as playbooks, allowing easier
code sharing with a-h-t and just in general being a better library for
talking ssh and executing commnads.
Closes: #1297
Approved by: jlebon
This is a logical followup to:
- "core: Add rpmostree.repo metadata to imported packages" AKA a52cb7d78e
- "compose: Add rpmostree.rpmmd-repo metadata to commits by default" AKA e7a42f70a9
Basically I'd like to display this in `status` at least, though how we render it
gets...interesting when there is layering in play. For now though let's just
capture the data the same way we do server side.
Theoretically we could reconstruct this data by walking all of the pkgrefs in
the tree but...ugh. It's just a lot simpler to have it in this form too just
like the server-side path.
Closes: #1296
Approved by: jlebon
Very often when I have a pending deployment, I just want to find out
what the diff is before rebooting. We do print it after whatever
transaction laid down the new deployment, though we may not reboot right
away. It seems like a common enough operation to warrant some shortcuts.
This makes `db diff` more useful by automatically subbing in the booted
deployment if less arguments are given:
rpm-ostree db diff
<diff booted and pending deployment>
rpm-ostree db diff $csum
<diff booted and $csum>
rpm-ostree db diff $csum1 $csum2
<as before, diff $csum1 and $csum2>
As before, we never try to even load the sysroot in the last invocation,
so it still works equally well on non-OSTree systems/unprivileged.
Closes: #1294
Approved by: cgwalters
Implemented in libostree in https://github.com/ostreedev/ostree/pull/1464
Let's display it - wrapping the command will come later.
I also just noticed `rpmostree_syscore_filter_deployments()` at least is
going to have to learn about pinning; will need to improve the test suite
around this too.
Closes: #1292
Approved by: jlebon
We had no coverage for `--check` and `--preview`. Now that they use the
more efficient auto-updates API, it should be fine to document/recommend
to users. Let's add some coverage in the existing auto-updates tests to
make sure we don't regress on them e.g. only working when auto-update
"check" mode is on.
Closes: #1268
Approved by: cgwalters
Rather than recalculating `cached-update` as part of transaction
cleanups and RpmostreedOS internal reloads, write it directly to a file
from `deploy_transaction_execute`. This gives two major benefits:
1. Auto-updates now has virtually zero impact to daemon startup time.
2. We get to directly use the `DnfSack` created during metadata refresh
rather than reconstructing it later on. This greatly simplifies code.
This makes use of new APIs in libdnf to skip filelists and load
updateinfo metadata right from the start.
Closes: #1268
Approved by: cgwalters
We would error out when trying to start the transient httpd service if
it already exists, e.g. from a previous test.
Depending on how we exit, the `vm_stop_httpd` trap for the previous test
might not have been able to kick in. I think this happens when we exit
using `fatal`, which just does an `exit 1`. It's not strictly an error,
so doesn't trip the `ERR` handler.
Let's just go the extra mile and explicitly delete transient services if
they already exist.
Closes: #1284
Approved by: cgwalters
This renames the remaining C files, tests, etc. There are only
a few hits for `jigdo` left; changing them would be a format break,
so let's wait to do that until we need to.
Closes: #1279
Approved by: jlebon
We're continuing an incremental renaming process; previously we changed
the most user-visible strings. Now we're doing some internal variables,
and notably the cached refs and the origin files - the latter set is
things that end up on disk.
This leaves the biggest items; renaming APIs, files, and tests.
Closes: #1276
Approved by: jlebon
Let's try to match expectations a bit from the dnf/yum world and
describe e.g. a `noarch` package become archful as an upgrade rather
than a removal and an addition. This was originally implicitly supported
(before PR #1230) by the fact that we didn't compare arches at all (and
in fact, arches don't even show up in a `db diff` output for modified
packages).
We bring this back here, but only in the simple case that it's a single
package. We still don't try to do any fancy handling for packages of the
same name.
Closes: #1272Closes: #1274
Approved by: cgwalters
I noticed that Ubuntu also uses the original "jigdo", so let's start
pulling off the band-aid here and do a mass rename.
For this first pass I'm focusing on CLI entrypoints and docs, as that's what
people are going to see; renaming all of the internal C functions, structure
variables etc. can come later.
Closes: #1269
Approved by: jlebon
This is an essential functionality rather than a nicety. Some
replacements can *only* be done without conflicts if we can remove
packages at the same time.
I do like that this has to be done explicitly, though OTOH, I can
definitely see folks wanting an `--allow-removals` type of switch in the
future.
Closes: #1255Closes: #1257
Approved by: cgwalters
Changes in a server-side tree can cause the need for clients to import different
objects from packages. For example, turning on documentation. Another more
subtle case is where an object might "move" from package A to B by being deleted
from A - then the jigdo build process will pick the B version.
We need a "cache validation key"; a way for the server to tell the client that
the objects it should import from the package have changed. Initially I was
thinking of using the libostree "content hash" but that would be awkward as we'd
have to do an import on the server side too.
After more consideration I realized a simple *count* of the number of objects
actually works, because (as I note in a comment) changing a file in the tree
will result in it ending up in the jigdoRPM (and count as a deletion). And
obviously adding or removing objects changes the count too.
In fact we could have done this *without* breaking the format by just having the
client start recording the number of xattr entries, but this adds greater
flexibility down the line since we can in theory change how we do cache
invalidation if we *really* need to (but at the cost of triggering clients to
redownload packages).
Note the client logic got moved around as now we need to parse all the xattrs
before we decide what packages to download.
My test case here is turning on docs - I noticed this actually affects *every*
package which was surprising to me; I expected at least some packages wouldn't
have docs. I'll double check this.
It'd be good to have a "moving object" case too which I may look at.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1197Closes: #1256
Approved by: jlebon
Since we changed things to have `jigdoSet = pkgSet`, we can just require exactly
`${npkgs}` here on import, which is what we found from `db list`.
Closes: #1256
Approved by: jlebon
What's happened up till now is supporting `rojig://` in the same way as
`ostree://`. However, part of the high level goal here is to reduce
the need for system administrators to understand ostree.
This patch set starts to introduce some of the ideas for client-side
changes as part of jigdo ♲📦:
https://github.com/projectatomic/rpm-ostree/issues/1081#issuecomment-348540604
Concretely, we start using `${repo}:${nevra}` instead of `rojig://`.
(v2): Keep `Version` (plus timestamp) as a split out field for maximum visual aid.
Also, let's be opinionated here and entirely drop the `Commit` checksum by
default. I believe the Cockpit guys were right here - versions are for humans.
The fact that we have a checksum is powerful; and we still show it with `status
-v`. The way I think of it is: the checksum shows we're really an image system.
But we don't need to show it by default.
Closes: #1240
Approved by: jlebon
Pick up security advisories when checking for pending updates and
include them in the `cached-update` property. On the client-side,
display them in the output of `status`.
This was part of the original vision for how useful a smart `check` mode
could be. It directly impacts how one manages their individual system
(e.g. when to reboot), and paves the way for integration into
higher-level apps that act at the cluster level.
Closes: #1249
Approved by: cgwalters
Fix logic to make sure we check if the refspec is of type `ostree://`
even when it's explicitly specified. Also fix `Deploy` in the case where
we didn't just `Download` the RPM diff by adding a new @checksum
parameter to the higher-level API.
Finally, add a basic test for the `GetCached*RpmDiff` APIs so we have at
least *some* coverage. This is also good prep for making sure we don't
break anything when we convert those APIs to use the more efficient
pkglist metadata. The tests completely ignore the `DownloadRpmDiff`
paths for now though.
Closes: #1250Closes: #1253
Approved by: cgwalters
This fleshes out an important piece of the story, showing that
we can support history versioning the same way that we did with
ostree.
Also it's very useful for testing; I'm going to extend the suite after this to
deploy the previous version, clean everything up, then upgrade and verify we
only download changed RPMs.
Closes: #1232
Approved by: jlebon
The use case here is to run our tests against the shipped tree;
this could be used by a-h-t for example:
https://github.com/projectatomic/atomic-host-tests/issues/74
I tried this with just `TESTS=basic` and it failed for `usroverlay`. So we'd
have to start adding feature detection to the test suite to make this truly
useful, but let's at least start with the basic bits now to play with it.
Closes: #1251
Approved by: jlebon
The goal here is to support `override replace kernel.x86_64.rpm`.
There's a whole lot of logic in libdnf to support having multiple
kernel packages installed. AIUI, that was implemented because:
1) The kernel is the biggest source of regressions
2) It'd be quite noticeable if all of your kernel modules were deleted
Of course point 2) applies to a lot of userspace too...that's something
rpm-ostree fixes of course.
Anyways, in some testing all we need to do really is just turn that
logic off unconditionally.
Closes: https://github.com/projectatomic/rpm-ostree/issues/946Closes: #1228
Approved by: jlebon
While debugging the recent `BDB0087 DB_RUNRECOVERY` issues that cropped
up recently, I came upon the fact that we're leaving leftover rpmdb
files in the rootfs on client-side assemblies. Let's clean those up too.
We do this in the `commit` phase, *after* the final time we access the
rpmdb to generate the pkglist.
Closes: #1247
Approved by: cgwalters
The way we import packages in jigdo mode is different from package layering; we
may only import a subset of files for example. In general, we need to treat
jigdo differently.
Related: https://github.com/projectatomic/rpm-ostree/issues/1197Closes: #1238
Approved by: jlebon
I *think* this is why our tests started failing recently. It seems somehow very
unlikely to me though that we'd somehow managed to avoid `f` in the boot
checksums until now, but without doing some math...it seems plausible.
Closes: #1243
Approved by: jlebon
I saw kalev's slides reference `rpm-ostree unlock`; this patch makes it exist.
In general, people have a hard time (understandably) grasping the distinction
between ostree and rpm-ostree; along with the goal of making ostree really
"libostree", let's start wrapping more commands where it makes sense.
I also took this opportunity to have a more descriptive name; it's important
to note that it *doesn't* overlay `/etc`, `/var`, or `/boot` for example.
Closes: #1233
Approved by: jlebon
Not all files from an RPM are necessarily removed during pkg erasure.
For example, files which are shared between pkgs shouldn't be deleted.
Similarly, not all files in an RPM are necessarily copied during pkg
installs. This is the case for multilib handling, which is a mess in its
own right. But such is the cost of trying to replace major parts of a
long-standing foundational project like RPM.
This patch adds some smarts to the way we do overlays and overrides to
handle these cases by calculating beforehand which files we *should't*
checkout/delete.
Closes: #1217Closes: #1145Closes: #1227
Approved by: cgwalters
Our complicated heuristics for handling multiple packages of the same
name comes back to bite us. In the multilib case, we can have packages
of the same NEVR, but different arch, sitting in the same tree.
Previously, even if the arch was different, we would still mark it as an
upgrade or downgrade. But that complicates things in the case of
multiple packages of the same name in the same tree.
We greatly simplify things here by making the diff algorithm dumber. We
now only consider a package as "modified" (i.e. upgraded/downgraded) if
it has the same NA (but different EVR). This makes handling multilib
cases natural and seems worth it overall vs trying to handle the odd
e.g. noarch <--> archful pkg transitions that could happen.
Closes: #1230
Approved by: cgwalters
This is an initial drop of support for:
`rpm-ostree rebase rojig://fahc:fedora-atomic-host`. We also
then support `rpm-ostree upgrade` from that.
There's a lot that could be improved here; the test coverage is relatively
minimal. A blocking issue there is having a realistic jigdo setup, and that's
going to require changing how we do testing. For now, this means that if we want
to e.g. change the format we'll have to temporarily disable this test, get the
format change in, update FAHC, then re-enable the test.
Closes: #1166
Approved by: jlebon
I'm working on supporting `override replace ./kernel-*.rpm`:
https://github.com/projectatomic/rpm-ostree/issues/946
But after battling the "installonly" logic in libdnf, I was confused why we
still had the files in `/usr/lib/modules`. It turned out to be because we only
remove files in `/usr`, but the code didn't handle UsrMove compat links.
There are a variety of approaches to fix this. Obviously a lot
of things get nicer in jigdo mode as we do UsrMove canonicalization
on import, and we could change this code to walk the imported pkg
ostree ref.
Another approach would be to walk the initial symlink, and check whether or not
it's the same as `/usr` (i.e. let the kernel do it).
For now though, what I chose to do was to scan the rootfs and find the UsrMove
compat links (i.e. we avoid hardcoding them again here). This is
fewer syscalls than the above and works well in practice.
Closes: #1226
Approved by: jlebon
This is what a lot of other tools do. It can get very verbose, with a
potentially huge amount of output if things are trickling in. This way
we're at least more friendly to someone running `cmd: rpm-ostree upgrade`
via Ansible or equivalent.
The slight hack here is that we *do* output `100%` on non-ttys to ensure we
print the result of the task.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1183Closes: #1225
Approved by: jlebon
This patch introduces a new `AutomaticUpdatePolicy` configuration. This
was a long time coming for rpm-ostree, given that its update model makes
it extremely apt for such a feature.
The config supports a `check` mode, which should be very useful to
Atomic Workstation users, as well as a `reboot` mode, which could be
used in its present form in simple single node Atomic Host situations.
There is still a lot of work to be done, including integrating
advisories, and supporting a `deploy` mode. This feature hopefully will
be leveraged as well by higher-level projects like GNOME Software and
Cockpit.
Closes: #1147
Approved by: cgwalters
In #875 AKA b46fc35901 we
added support for the `releasever` option in treefiles. I am
pretty sure it worked at the time...but I didn't add tests.
Either it never worked or some refactoring broke it. The whole chain of
`GKeyFile` → `GVariant` is so confusing. Anyways fix it by copying the string.
Now let's use it by default in the compose tests, and while we're here bump
those to F27.
I'm doing this patch now as I was playing with doing a compose from
the `/usr/share/rpm-ostree/treefile.json` and wanted to use the stock
`.repo` files.
Closes: #1220
Approved by: jlebon
It obviously shouldn't block the ability to install, and anyways
the single use in `ksh` is not compelling. If someone comes up with
one we can revisit supporting this.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1216Closes: #1218
Approved by: jlebon
When I tried to use my WIP client patches to do:
`rpm-ostree rebase rojig://fahc:fedora-atomic-host`,
I got a missing file object which turned out to
be the client importing the i686 RPMs.
This was passing in the test suite because we don't mirror i686 of course, but
on the client side right now we end up using all enabled repos, and since Fedora
is multiarch, the behavior is going to be...not predictable.
Thinking a bit about on this problem I actually happened to recall
the RPM `%{_isa}` macro which is used in Fedora in various places;
for example to "arch bind" `-devel` packages to their base. See
for example [this case](33c7dc02bc/f/ostree.spec (_79)) in libostree.
As I noted at first, the core problem here is that the "final"
RPM architecture field is not symmetric in any way with the definition
of that `%{_isa}` macro. See:
d9d47e0114/installplatform (L25)
The *third* solution I ended up on here is to iterate over the
`Provides` on the server side and we take the first thing
that matches `Provides: %{name}(whatever)`.
I briefly thought about trying to somehow drive into libsolv the
logic to prefer the jigdoRPM's native architecture...IIRC yum did
something like that in the past but it was never done in libsolv?
Anyways the dependencies here are now more correct, so other tools
will also handle it.
Closes: #1213
Approved by: jlebon
For CI purposes at least, these features are already part of FAH and the
latest CAH smoketested. Anyway, we're already hacking the latest ostree
into place for `--keep-metadata`. If it already supports the latter,
then clearly it'll support those.
Closes: #1215
Approved by: cgwalters
This reverts commit 1ef259ed76.
Basically, for historical POSIX compatibility, `errexit` doesn't work at
all inside functions called from an if-statement/boolean context. This
is something I had already learned (and forgotten) when hacking on PAPR.
There are ways around this, but it's just not worth the added complex
shell goop for what it gives us.
http://mywiki.wooledge.org/BashFAQ/105Closes: #1215
Approved by: cgwalters
I spent some time trying to figure out why on CentOS my changes to
`/etc/rpm-ostreed.conf` didn't seem to have any effect. It turns out
that for some reason I didn't spend too much time looking into, the
autoconf in CentOS defaults to `${prefix}/etc` when `--prefix` is given.
This is also why I was under the erroneous assumption that D-Bus config
files were in `/usr/etc` in CentOS. This patch fixes that as well.
Closes: #1215
Approved by: cgwalters
Let's standardize on the default config file when running tests. We copy
the original out of the way and install the default one so tests can do
whatever they want with it.
This also strengthens the post-test cleanup to make sure we rebase back
to the local vmcheck branch, in case we're somehow on a different branch
with the exact same commit.
Closes: #1212
Approved by: cgwalters
The output from `rpmbuild` makes test results harder to comb through
when debugging. Let's just dump all that to file and only output it if
something goes wrong.
Closes: #1212
Approved by: cgwalters
This fixes an asymmetry we introduced when rendering using the `ostree://`
prefix; we now support parsing it as well. We reject the `rojig://` prefix
explicitly for now.
On the implementation side...things are messy. My thought is that we do our best
to canonicalize at the entrypoints for both the client side as well as the
server side. However, for backwards compatibility we can't go all the way to
writing out `ostree://` to disk in the origin files.
I'm actually uncertain if Cockpit will deal with this...need to test that.
Closes: #1210
Approved by: jlebon
One issue right now is that repeated applications of `make vmoverlay`
will result in losing the source title.
Let's use the new `--keep-metadata` in subsequent `make vmoverlay`
invocations to make sure it gets transferred over correctly.
Closes: #1205
Approved by: cgwalters
This makes it much easier to know how old the base we're working from is
and whether it's time to upgrade before hacking some more.
Closes: #1205
Approved by: cgwalters
This is essentially a revert of #1190. I initially changed from storing
the epoch, version, and release separately to together to avoid endian
issues with epoch. This works around that instead by just encoding it as
a string.
As is done in our custom NEVRA printer (which before #1190 we used
here), we make sure that epochs of "0" are not printed as part of the
`nevra` or `evr` members.
Closes: #1198
Approved by: cgwalters
Basically we're choosing the libdnf behavior of omitting `Epoch: 0` which IMO is
the sane thing to do even though in *theory* they're different RPMs but
who would ever ship both without incrementing release/etc?
More information in: https://github.com/projectatomic/rpm-ostree/pull/187Closes: #1198
Approved by: cgwalters
`grep` supports checking multiple fixed strings separated by newlines,
but it's mostly just easier to pass them as separate arguments, so let's
support that. This is now at parity with the similar
`assert_file_has_content`.
Will upstream this to ostree as well once reviewed.
Closes: #1200
Approved by: cgwalters
Following up on `--ex-jigdo-output-rpm`, add support for writing the entire set
to an output directory. This is intended for use cases like FAHC, where we're
generating data outside of the upstream Fedora infrastructure. Further, we want
to support having our own history stored reliably, even if upstream prunes RPMs.
Now, this can be interesting even for upstreams like Fedora, as it naturally
captures just the subset of RPMs; doing full history support for that would
likely be a lot more palatable than for Everything.
Closes: #1165
Approved by: jlebon
Part of the goal of jigdo ♲📦 is to support organizations switching to *only*
providing RPMs. An intermediate step there is to "lock" the repo and jigdo
together; we don't want to update the ref if building the jigdoRPM fails.
Add an option to perform `rpm-ostree compose tree` and `rpm-ostree ex
commit2jigdo` together; notably we generate a commit, but only update the ref
once the jigdoRPM is built.
Closes: #1165
Approved by: jlebon
- Actually use separate `${test_tmpdir}` for test setup (closes a race)
- Merge stdout/stderr (more readable)
- Ensure logs are renamed to `.txt` even on failure
- Use `--progress` for some feedback
- Use `-j +1` so that even on unicore machines we get at least 2
jobs (and in general NCPUS+1)
Closes: #1188
Approved by: jlebon
Basically the `rpmostree_context_relabel()` call we had in the treecompose path
for unified core didn't actually have any effect as the core code did a relabel
and unset the array.
I think this may actually be a regression from: https://github.com/projectatomic/rpm-ostree/pull/1137
though I didn't verify.
Anyways looking at this, the code is a lot simpler if we change the API so that
the "normal" relabeling is folded into `rpmostree_context_assemble()`. Then we
change the public relabel API to be "force relabel" which we use in the unified
core 🌐 treecompose path.
This shrinks the jigdoRPM for FAH from 90MB to 68MB.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1172Closes: #1173
Approved by: jlebon
This fixes a large swath of compatibility issues, for the same reasons as
overlayfs makes a lot of things Just Work. The ugly part of course is
doing hidden copyups inside the filesystem.
We've gone quite a long time with the "pure rofiles" mode, and have made changes
to various bits of userspace to be compatible with it. But what finally made me
give up on that is glibc's locale-archive; there's a patch for it that
is stalled, but even if it was applied we would still need to work with
older glibc.
This issue comes to the fore in unified core 🌐 mode, as without this
we won't get a correct locale archive.
Closes: #1171
Approved by: jlebon
Over a year later, the "opening the host rpmdb" bug is fixed,
so we can do composes in parallel ∥, hooray!
I'm dusting this off since we were running into CI (PAPR) timeouts
when I was adding more to the compose tests.
Closes: #545
Approved by: jlebon
This is the first step towards unifying how we introspect packages from
a specific commit. We currently do this in three ways: libdnf, librpm,
and now `rpmostree.rpmdb.pkglist`. I'd like to get to a point where we
only have `rpmostree.rpmdb.pkglist` and libdnf, the latter only when
more complex queries are required.
This patch teaches the `db diff` command to make use of the new db diff
API so that it can work even on metadata-only commits. This is relevant
for use cases mentioned in #558.
I didn't get rid of the `rpmhdrs_diff` functions right now because of
the `--changelogs` option: libdnf currently does not expose this, so we
fall back to the previous API in that case. OTOH, I wonder how much it's
actually used in the wild; maybe we could just nix it?
Closes: #1162
Approved by: cgwalters
This is used in critical paths like pungi, so let's be sure it works;
the semantics are a bit subtle as it overrides setting the ref.
Closes: #1161
Approved by: jlebon
To make the new pkglist metadata even more usable, let's insert it
sorted. This ensures that we can bsearch the GVariant on the
client-side.
Enhance the bsearch utility function we have to deal with duplicate key
names. Although this is not the case today in rpm-ostree-managed streams
I know, some packages are allowed to have multiple versions installed,
so let's make sure we handle that deterministically by always returning
the first (earliest) version of the package.
Closes: #1158
Approved by: cgwalters
The new `rpmdb.pkglist` metadata is a cheap way of retrieving the set of
packages in a commit. I'd like to make use of it as much as possible vs.
checking out the rpmdb and setting up a DnfSack.
Of course, in the case of layered commits, it doesn't matter *as* much,
because a layered commit being present in the repo should mean that a
deployment is currently using it, and we should learn to reuse the rpmdb
checkout of that deployment. Though keeping it consistent across both
server and client commits makes implementing `OstreeDeployment`-agnostic
things like `db diff` more efficient too. I also plan to use this in the
upcoming auto-update code.
Closes: #1158
Approved by: cgwalters
This fixes another thing broken with `compose --ex-unified-core`;
for e.g. `/usr/bin/ping` from `iputils`, the classic example of a filecaps
binary.
As I'm writing this commit message I realize it will actually also
take effect for package layering unnecessarily; we'll pointlessly
break the hardlink. But eh, it doesn't matter right now, we can
optimize that later.
Closes: #1151
Approved by: jlebon
Rather than entirely symlinking `systemctl` → `/bin/true`, in order
to e.g. have NetworkManager be enabled, we need to process presets.
This is one of the things that's breaking FAHC where I did a
`--ex-unified-core` deployment.
(Actually it's a bit tempting to run a mass preset pass at the end,
but for now let's do this)
Implementation note: this is our first use of GResources, which
is a handy way to embed data into our final binary.
Closes: https://github.com/projectatomic/rpm-ostree/issues/550Closes: #1148
Approved by: jlebon
This rolls up several libglnx changes: https://github.com/GNOME/libglnx/pull/101
Now of course things are trickier here because we have an internal
abstraction over directly emitting to a console versus sending the
result over DBus. Further complicating things is that some things
call into libdnf and thus *require* use of `DnfState` which does
not give us the "n items" information, versus other parts which
we implement and can do what we want.
Even *further* complicating things is that we have to take care around non-CLI
callers like Cockpit; so I didn't try to pass the "n items" over DBus, rather
just reimplemented the "insert into text" that libglnx is doing.
Anyways overall this looks better IMO for all cases.
Update submodule: libglnx
Closes: #1143
Approved by: jlebon
This is prep for a rework of
https://github.com/projectatomic/rpm-ostree/pull/621
For a no-op `rpm-ostree upgrade` (i.e. no updates available), as long as
layering is enabled, we pay the cost of checking out the base tree, *mostly*
only to get the base rpmdb.
This is prep for fixing that down the line by knowing we always have the "base"
tree's rpmdb checked out. Then in the layering case we only modify
`/usr/share/rpm` (eventually that will point to `/usr/lib/sysimage/rpm`).
Teaching `rpmostree-core.c` about this can follow on later.
Closes: #1142
Approved by: jlebon
Having the "jigdo set" in repodata makes it so we can parallel download the
jigdo RPM with the set. However for now, I kept the jigdo set in the jigdoRPM,
since that way it'll be covered by the signature.
Also, this changes the way we inject metadata to use a magic comment string,
since trying to pass a gigantic macro to `rpmbuild` via its argv didn't work out
so well (it looks like rpmbuild eats newlines). This approach is more robust.
Closes: https://github.com/projectatomic/rpm-ostree/issues/1132Closes: #1140
Approved by: jlebon
With unified core 🌐 we are *relying* on the devino cache
for correctness when using `bare-user` repos. Otherwise lots
of bad things will happen as we won't hit the happy path from
[this libostree PR](https://github.com/ostreedev/ostree/pull/1297)
(I should probably add an assertion there that we aren't trying to commit
`user.ostreemeta`).
It looks like I had this working in some of the old unified core WIP patches,
but it was lost when rebasing 🏄.
We noticed this when I was trying to deploy jigdo in FAHC and the system
wouldn't boot as various things rely on those suid transitions.
Closes: #1139
Approved by: jlebon
We don't want to have to download all of `/usr/share/rpm` just to get
the list of packages used to compose the tree. This is fundamental
information that needs to be easier to discover. So let's stick it right
in the commit metadata. There's various use cases for this information,
including easily checking for and displaying updates and a pkglist-aware
version of `ostree log`.
Closes: #1134
Approved by: cgwalters
This function is now basically reimplementing all of the intelligence
we have in libostree today. It wouldn't have worked before we
fixed https://github.com/ostreedev/ostree/issues/1165 but now
that's done, conceptually we just need to call
`ostree_repo_commit_modifier_set_sepolicy()`. However, I had
to drop the "number of files changed" since currently the libostree
API doesn't support that.
Also, in order to detect the case that content changes at all (so
we still have some useful information in the journal), implement
a "content hash" check.
See also https://github.com/projectatomic/rpm-ostree/pull/1123
(And note we *don't* use `DEVINO_CANONICAL` here)
Closes: #1138
Approved by: jlebon
Right now the fact that one can only cancel via `Ctrl-C` of an existing client
process is rather frustrating if for example one's ssh connection to a machine
drops. Now, upon reconnecting, one can easily `rpm-ostree cancel` a hung update
or whatever rather than doing the more forcible `systemctl stop rpm-ostreed`
(which is safe of course, unless livefs is involved).
Closes: #1019
Approved by: jlebon
Now that the importer *only* imports into OSTree repos, let's
clean up the API so that the `OstreeRepo` and `OstreeSePolicy`
are passed as constructor args.
Also rework things so there's only one constructor API that
steals the fd.
This is prep for adding another async import API.
Closes: #1124
Approved by: jlebon
We are going to want versioning on the jigdo RPMs, since it's
highly likely things change.
This is done via new magic '-D rpmostree_jigdo_meta` macro, which we can also
use for other things down the line.
Closes: #1135
Approved by: jlebon
Let's "repo bind" the OIRPM by default; this makes the rpm-md repo feel a lot
more like an ostree remote, and IMO is just a really good idea in general to
increase predictabilty.
Closes: #1130
Approved by: jlebon
Note this PR requires [bubblewrap 0.2.0](https://github.com/projectatomic/bubblewrap/releases/tag/v0.2.0).
Change our bwrap invocations drop truly dangerous capabilities like
`cap_sys_admin` and `cap_sys_module` just like Docker does today. Because of the
popularity of Docker, we can be pretty sure that most RPM scripts should have
adapted to this (although a problematic area here is that traditional librpm
doesn't actually error out if scripts fail).
There are two reasons to do this:
- We want "offline" updates by default; updates shouldn't affect the
running system. If we prepare the new root in the background, a
%post shouldn't restart a service for example. We already "handle"
this by making `systemctl` a symlink to `/bin/true`, but this approach
also shuts off `%post`s that do e.g. `insmod`.
- Protection against accidental system damage
Closes: #1099
Approved by: jlebon
Tracking issue: https://github.com/projectatomic/rpm-ostree/issues/1081
To briefly recap: Let's experiment with doing ostree-in-RPM, basically the
"compose" process injects additional data (SELinux labels for example) in an
"ostree image" RPM, like `fedora-atomic-host-27.8-1.x86_64.rpm`. That "ostree
image" RPM will contain the OSTree commit+metadata, and tell us what RPMs we
need need to download. For updates, like `yum update` we only download changed
RPMs, plus the new "oirpm". But SELinux labeling, depsolving, etc. are still
done server side, and we still have a reliable OSTree commit checksum.
This is a lot like [Jigdo](http://atterer.org/jigdo/)
Here we fully demonstrate the concept working end-to-end; we use the
"traditional" `compose tree` to commit a bunch of RPMs to an OSTree repo, which
has a checksum, version etc. Then the new `ex commit2jigdo` generates the
"oirpm". This is the "server side" operation. Next simulating the client side,
`jigdo2commit` takes the OIRPM and uses it and downloads the "jigdo set" RPMs,
fully regenerating *bit for bit* the final OSTree commit.
If you want to play with this, I'd take a look at the `test-jigdo.sh`; from
there you can find other useful bits like the example `fedora-atomic-host.spec`
file (though the canonical copy of this will likely land in the
[fedora-atomic](http://pagure.io/fedora-atomic) manifest git repo.
Closes: #1103
Approved by: jlebon
`test -n` would always be true. Also while I have the patient open
let's make writing the file atomic. Maybe someday I'll be motivated
enough to write an `O_TMPFILE` patch for bash.
Closes: #1103
Approved by: jlebon
When we added the `--ex-unified-core` option our caching story got
very messy because the non-unified core caches RPMs, but unified
does ostree repo caching.
For jigdo, we want the RPMs. Fix this by mirroring the RPMs using
`--download-only` and pointing the tests consistently at that.
Closes: #1122
Approved by: jlebon
We originally needed the pkgcache to be a separate repo due to ostree's
overzealous pruning policies. The idea was to maintain multiple commits
in each pkg branch for different SELinux policies. In practice, there's
not much use in maintaining old copies and it's just easier to always
relabel on the fly. So then, the need for a separate repo completely
melts away.
This helps simplify the mental model a bit and allows us to avoid subtle
issues like #1047. Note however that the core is still capable of
handling split repos for the `--ex-unified-core` compose use case. Once
that and the jigdo work are a bit more settled, we can have a clearer
picture of how to simplify the core further.
The tricky bit is migrating the cache. When deploying, we check if a
pkgcache repo exists and migrate its refs if so. We then leave behind a
symlink to the system repo to remain compatible with older rpm-ostrees.
Closes: #1055
Approved by: cgwalters
Right now each ostree txn incurs a `syncfs()`; see
https://github.com/ostreedev/ostree/issues/1184
And before this patch, we were doing a txn per package import.
We can really do better in libostree - we'll fix that, but in the short term
let's use a bigger txn for every package. However, the obvious change here of
simply hoisting up the txn is that on failure for imports, we'd discard all
downloaded packages. We fix that by changing the auto txn API to have
a `commit_on_failure` boolean, and use it in cases where we're doing
imports.
This is prep work for jigdo, where we'll be using the import path all the time.
My bigger plan is to do multithreaded imports.
Closes: #1116
Approved by: jlebon
It took me way, way, way too long to debug that my dev container somehow missing
`ostree-grub2` caused the script to fail early, but we'd still continue trying
to sync, leading to us not actually changing the installed `rpm-ostree` binary,
leading to test suite failures.
Also add the chmod hammer at the top so we can delete even if we fail after
doing an rsync.
Closes: #1115
Approved by: jlebon
In the jigdo path we don't actually want to import the OIRPM literally
into ostree. I considered adding jigdo logic into `rpmostree-unpacker.c`
but it'd be a mess as the functionality is quite logically separate
from importing.
So split off an `unpacker-core.c` file which has the bare libarchive+RPM
helpers, and rename `RpmOstreeUnpacker` to `RpmOstreeImporter`.
Closes: #1110
Approved by: jlebon
We removed this in review, but I rediscovered why I added it. We
fail the `g_assert (sepolicy_matches)` if we already had packages
done with the final label when we go to reuse the cache.
(Basically, if we use the cache multiple times it's hard to avoid
relabeling all the time which is unfortunate...gets back a bit to
a way to annotate pkgcache commits as supporting multiple policies)
Closes: #1109
Approved by: jlebon
The "--ex" prefix here means it's an experimental option. A tremendous change
here is that start to support non-uid 0, but there are various things to fix there;
the unpacker for example needs to learn to set imported objects fully based
on the rpmfi information (i.e. default to uid 0, since libarchive gives the
current uid by default).
And even when run as uid 0, there are some bugs, though I'm not sure
of any showstoppers yet. For example, dracut's `dracut-install` calls
`cp --preserve=xattrs` which fails to copy the `user.ostreemeta` xattrs
from a checkout (it shouldn't be copying that anyways...)
Nevertheless, the infrastructure behind this really helps (is almost a hard
requirement for) the [jigdo effort](https://github.com/projectatomic/rpm-ostree/issues/1081).
Which is really only true due to SELinux - we need to import the packages,
then generate the final tree to get the final policy, then use that policy
to relabel all of the packages.
Closes: #940
Approved by: jlebon
In an unprivileged case, we can't do this on the real filesystem. For
`ex container`, we want to completely ignore uid/gid. I added a test
installing `httpd` which failed previously.
TODO: For non-root `--ex-unified-core` we need to do it as a commit modifier.
Closes: #940
Approved by: jlebon
More work for unified core; this was one of the biggest parts of the `ostree
diff` when comparing a unified core treecompose of Fedora Atomic Host versus
the libdnf path.
This is also the first case where we're teaching the pkgcache to redownload
based on state.
Closes: #1105
Approved by: jlebon
I was trying to figure out why:
rpm-ostree compose tree --repo repo/ manifest.json
would result in:
error: opendir(manifest.json): No such file or directory
It turned out to be because we had `--repo` in *both* the `install`
options and the `commit` options. This makes sense since both these
subcommands need to be given a repo. However, in the `tree` case, we
were adding both arrays, which meant we inherited two `GOptionEntry`s
for `--repo`. This confused glib2 which consumed not one but two CLI
arguments when looking for the argument associated with `--repo`.
Our CI didn't notice this because it uses the `--repo=foo` notation,
which doesn't throw off glib2.
Fix this by factoring out the `--repo` option into a separate array so
that in the `tree` case, it only gets added once. Exercise the fix in CI
by using the two argument notation for `--repo`.
Closes: #1101
Approved by: cgwalters
We have a *lot* of experimental functionality. I think the
`override` bits are fleshed out enough now that we can lift
the `ex` designation. For example, jlebon fixed SELinux
labeling in the presence of override-replace.
Closes: #1089
Approved by: jlebon
Added unit tests for rpm-ostree ex kargs --delete,
--append and --replace.
Also exposed two getter functions for kargs table
and array so people can retrieve information from
kargs.
Also includes a minor fix for a bug caught by the unit
test.
Closes: #1013
Approved by: cgwalters
Refined a bit for the previous written kargs test.( mainly for
checking kargs after rollback )
Added tests for import-proc-cmdline and deploy-index option
Closes: #1013
Approved by: cgwalters
Refactor the similar code portion from delete option into
a new function, and reuse it in the replace logic.
Add the replace functionality to allow kernel arguments
replacement. Now support two different types of replacement:
1: replace by key=value when there is only one key existing
e.g we have 'hi=hello' as our one of the arguments (hi only has
one value associated with it) , and 'rpm-ostree ex kargs --replace hi=new'
will replace the value and change hi=hello into hi=new.
2: replace by key=value=new_value for all other replacements
( this will work for swapping single value pair too !)
Some tests for rpm-ostree ex kargs --replace are added
for future regression.
Closes: #1013
Approved by: cgwalters
