wip: partial cancel (#16)

* wip: partial cancel

The partial cancel request seems to work, but later when
I try to do Info or Cancel (the regular cancel) for the
job I get an error (I assume the job does not exist,
but because of the bug with the error messages not passing
forward I cannot see any further output for it).

* test: improve testing setup for match and cancel

Problem: the current testing is not standard for Go,
and makes it hard to understand or run in units.
Solution: move testing into proper test alongside
package, and break apart testing for cancel and match.
Additionally, we are using the same graphs / jobspecs
from flux-sched.

* extend tests to cancel job
* docker: update image bases

Problem: older bases do not have new enough gcc to build flux-sched.

Signed-off-by: vsoch <vsoch@users.noreply.github.com>

Author: vsoch

.github/workflows/build-deploy.yaml

@@ -20,7 +20,7 @@ jobs:
       uses: actions/checkout@v4
     - uses: actions/setup-go@v4
       with:
-        go-version: ^1.21
+        go-version: ^1.22
     - name: GHCR Login
       if: (github.event_name != 'pull_request')
       uses: docker/login-action@v2

.github/workflows/main.yaml

@@ -21,10 +21,10 @@ jobs:
       fail-fast: false
       matrix:
         # container base and lib prefix
-        test: [["fluxrm/flux-sched:jammy", "/usr/lib"],
-               ["fluxrm/flux-sched:fedora38", "/usr/lib64"],
+        test: [["fluxrm/flux-sched:noble", "/usr/lib"],
+               ["fluxrm/flux-sched:fedora40", "/usr/lib64"],
                ["fluxrm/flux-sched:bookworm-amd64", "/usr/lib"],
-               ["fluxrm/flux-sched:el8", "/usr/lib64"]]
+               ["fluxrm/flux-sched:el9", "/usr/lib64"]]
 
     container:
       image: ${{ matrix.test[0] }}
@@ -34,14 +34,15 @@ jobs:
     - name: Setup Go
       uses: actions/setup-go@v4
       with:
-        go-version: ^1.21
+        go-version: ^1.22
 
-    # TODO: we should consider distributing the header files with the release builds
     - name: flux-sched build
-      run: git clone https://github.com/flux-framework/flux-sched /opt/flux-sched    
-    - name: Build
-      run: LIB_PREFIX=${{ matrix.test[1] }} make build
-    - name: Test Binary
-      run: LIB_PREFIX=${{ matrix.test[1] }} make test-binary
-    - name: Test Modules
-      run: make test-modules
+      run: git clone https://github.com/flux-framework/flux-sched /opt/flux-sched
+#    - name: build flux-sched
+#      run: |
+#        here=$(pwd)
+#        cd /opt/flux-sched
+#        mkdir build && cd build && cmake ../ && make -j && sudo make install
+#        cd $here
+    - name: Test
+      run: LIB_PREFIX=${{ matrix.test[1] }} make test-v

Dockerfile

@@ -14,6 +14,9 @@ RUN wget https://go.dev/dl/go${GO_VERSION}.linux-amd64.tar.gz  && tar -xvf go${G
 ENV PATH=$PATH:/usr/local/go/bin:/home/vscode/go/bin
 
 RUN git clone https://github.com/flux-framework/flux-sched /opt/flux-sched
+RUN cd /opt/flux-sched && \
+    export FLUX_SCHED_VERSION=0.40.0 && \
+    mkdir build && cd build && cmake ../ && make -j && sudo make install
 
 # Assuming installing to /usr/local
 ENV LD_LIBRARY_PATH=/usr/lib:/usr/lib/flux:/usr/local/lib

Makefile

@@ -2,6 +2,7 @@ HERE ?= $(shell pwd)
 LOCALBIN ?= $(shell pwd)/bin
 JGF ?= $(HERE)/cmd/test/data/tiny.json
 JOBSPECS ?= $(HERE)/cmd/test/data/jobspecs
+CANCELDATA ?= $(HERE)/cmd/test/data/cancel
 
 # This assumes a build in the .devcontainer Dockerfile environment
 FLUX_SCHED_ROOT ?= /opt/flux-sched
@@ -17,43 +18,22 @@ LD_LIBRARY_PATH=$(LIB_PREFIX):$(LIB_PREFIX)/flux
 BUILDENVVAR=CGO_CFLAGS="-I${FLUX_SCHED_ROOT} -I${FLUX_SCHED_ROOT}/resource/reapi/bindings/c" CGO_LDFLAGS="-L${LIB_PREFIX} -L${LIB_PREFIX}/flux -L${FLUX_SCHED_ROOT}/resource/reapi/bindings -lreapi_cli -lflux-idset -lstdc++ -ljansson -lhwloc -lflux-hostlist -lboost_graph -lyaml-cpp"
 
 .PHONY: all
-all: build
+all: test
 
 .PHONY: test
-test: test-binary test-modules
-
-.PHONY: test-modules
-test-modules: 
-	go test -v ./pkg/types
-
-.PHONY: test-binary
-test-binary: 	
-	LD_LIBRARY_PATH=$(LD_LIBRARY_PATH) $(LOCALBIN)/test --jgf=$(JGF) --jobspec=$(JOBSPECS)/test001.yaml
-
-# test001_desc="match allocate 1 slot: 1 socket: 1 core (pol=default)"
-# test_expect_success "${test001_desc}" '
-#    ${main} --jgf=${jgf} --jobspec=${jobspec1} > 001.R.out &&
-#    sed -i -E "s/, 0\.[0-9]+//g" 001.R.out &&
-#    test_cmp 001.R.out ${exp_dir}/001.R.out
-#'
-
-#test002_desc="match allocate 2 slots: 2 sockets: 5 cores 1 gpu 6 memory"
-#test_expect_success "${test002_desc}" '
-#    ${main} --jgf=${jgf} --jobspec=${jobspec2} > 002.R.out &&
-#    sed -i -E "s/, 0\.[0-9]+//g" 002.R.out &&
-#    test_cmp 002.R.out ${exp_dir}/002.R.out
-#'
+test: 	
+#	$(COMMONENVVAR) $(BUILDENVVAR) LD_LIBRARY_PATH=$(LD_LIBRARY_PATH) go test -count 1 -run TestCancel -ldflags '-w' ./pkg/fluxcli ./pkg/types
+	$(COMMONENVVAR) $(BUILDENVVAR) LD_LIBRARY_PATH=$(LD_LIBRARY_PATH) go test -ldflags '-w' ./pkg/fluxcli ./pkg/types
+
+.PHONY: test-v
+test-v: 	
+#	$(COMMONENVVAR) $(BUILDENVVAR) LD_LIBRARY_PATH=$(LD_LIBRARY_PATH) go test -count 1 -run TestCancel -v -ldflags '-w' ./pkg/fluxcli ./pkg/types
+	$(COMMONENVVAR) $(BUILDENVVAR) LD_LIBRARY_PATH=$(LD_LIBRARY_PATH) go test -v -ldflags '-w' ./pkg/fluxcli ./pkg/types
 
 .PHONY: $(LOCALBIN)
 $(LOCALBIN):
 	mkdir -p $(LOCALBIN)
 
-# This serves as a single test file to build a dummy main to test
-.PHONY: build $(LOCALBIN)
-build: 
-	mkdir -p $(LOCALBIN)
-	$(COMMONENVVAR) $(BUILDENVVAR) go build -ldflags '-w' -o $(LOCALBIN)/test cmd/test/test.go
-
 .PHONY: clean
 clean:
 	rm -rf $(LOCALBIN)/test