|
1 | | -# Snakemake Executor Google Batch |
2 | | - |
3 | | -> This is currently a skeleton and not ready for use, but come back soon! 🎃️ |
| 1 | +# Snakemake executor plugin: google-batch |
4 | 2 |
|
5 | 3 | This is the [Google Batch](https://cloud.google.com/batch/docs/get-started) external executor plugin for snakemake. |
6 | | -If you are migrating from Google Life Sciences see [this documentation](https://cloud.google.com/batch/docs/migrate-to-batch-from-cloud-life-sciences). For the underlying Python SDK, see [google-cloud-batch](https://github.com/googleapis/google-cloud-python/tree/main/packages/google-cloud-batch) on GitHub. |
7 | | - |
8 | | -## Usage |
9 | | - |
10 | | -### Setup |
11 | | - |
12 | | -You'll likely want to start by setting up [application default credentials](https://cloud.google.com/docs/authentication/provide-credentials-adc#how-to) |
13 | | -The easiest thing to do is run: |
14 | | - |
15 | | -```bash |
16 | | -gcloud auth application-default login |
17 | | -``` |
18 | | - |
19 | | -### Quick Start |
20 | | - |
21 | | -The basic usage is, from a directory with your Snakefile, to ask for `googlebatch` as the |
22 | | -executor. |
23 | | - |
24 | | -```bash |
25 | | -$ snakemake --jobs 1 --executor googlebatch |
26 | | -``` |
27 | | - |
28 | | -You are minimally required to provide a project and region, and can do this through the environment or command line: |
29 | | - |
30 | | -```bash |
31 | | -export SNAKEMAKE_GOOGLEBATCH_PROJECT=myproject |
32 | | -export SNAKEMAKE_GOOGLEBATCH_REGION=us-central1 |
33 | | -snakemake --jobs 1 --executor googlebatch |
34 | | -``` |
35 | | - |
36 | | -or |
37 | | - |
38 | | -```bash |
39 | | -export SNAKEMAKE_GOOGLEBATCH_PROJECT=myproject |
40 | | -export SNAKEMAKE_GOOGLEBATCH_REGION=us-central1 |
41 | | -snakemake --jobs 1 --executor googlebatch --googlebatch-project myproject --googlebatch-region us-central1 |
42 | | -``` |
43 | | - |
44 | | -You can provide one or more custom arguments, as shown in the table below, to customize your batch run. |
45 | | -Note that batch offers setup snippets to help with more complex setups (e.g,. MPI). See [batch snippets](#batch-snippets) |
46 | | -for more information. |
47 | | - |
48 | | - |
49 | | -### Logging |
50 | | - |
51 | | -For logging, for an interactive run from the command line we provide status updates in the console you have running locally. For full logs, you can |
52 | | -go to the [Google Cloud Batch interface](https://console.cloud.google.com/batch/jobs?project=llnl-flux) and click |
53 | | -on your job of interest, and then the "Logs" tab. If you don't see logs, look in the "Events" tab, as usually there |
54 | | -is an error with your configuration (e.g., an unknown image or family). |
55 | | - |
56 | | -#### Isolated Logs |
57 | | - |
58 | | -If you need to retrieve logs for a job outside of this context (e.g., after a run or in a Pythonic test) you can use the provided script in [example](example). |
59 | | -Here is how to run it using the local poetry environment. You can either provide `--project` and `--region` or export the environment variables for them |
60 | | -described above. |
61 | | - |
62 | | -```bash |
63 | | -# <jobid> |
64 | | -poetry run python example/show-logs.py a-898674 |
65 | | -``` |
66 | | - |
67 | | -Note that this is currently provided as a helper script because the [Google Cloud API limits](https://cloud.google.com/logging/quotas#api-limits) |
68 | | -set a rate limit of 60/minute. |
69 | | - |
70 | | -> Number of entries.list requests 60 per minute, per Google Cloud project |
71 | | -
|
72 | | -For some perspective, a "hello world" job will produce over 3K lines of logs, and (without a sleep between calls) |
73 | | -the ratelimit is hit very easily. We are currently assessing strategies to deliver full logs to .snakemake logging files |
74 | | -without hitting issues with this rate limit. It looks possible to create "[sinks](https://cloud.google.com/logging/docs/routing/overview#sinks)" using |
75 | | -Pub Sub, however this would be adding an extra API dependency (and cost). |
76 | | - |
77 | | -### Arguments |
78 | | - |
79 | | -And custom arguments can be any of the following, either on the command line or provided in the environment. |
80 | | - |
81 | | -| Name | Description | Flag | Type | Environment Variable | Required | Default | |
82 | | -|------|-------------|------|------|----------------------|----------|---------| |
83 | | -| project | The name of the Google Project | `--googlebatch-project` | str | `SNAKEMAKE_GOOGLEBATCH_PROJECT` | True | unset | |
84 | | -| region | The name of the Google Project region (e.g., us-central1) | str | `--googlebatch-region` |`SNAKEMAKE_GOOGLEBATCH_REGION` | True | unset | |
85 | | -| machine_type | Google Cloud machine type or VM (mpitune configurations are on c2 and c2d family) | str | `--googlebatch-machine-type` | | False | c2-standard-4 | |
86 | | -| image_family | Google Cloud image family (defaults to hpc-centos-7) | `--googlebatch-image-family` | str | | False | hpc-centos-7 | |
87 | | -| image_project | The project the selected image belongs to (defaults to cloud-hpc-image-public) | `--googlebatch-image-project` | str | | False | cloud-hpc-image-public | |
88 | | -| bucket | A bucket to mount with snakemake data | `--googlebatch-bucket` | str | `SNAKEMAKE_GOOGLEBATCH_BUCKET` | True | unset | |
89 | | -| mount_path | The mount path for a bucket (if provided) | `--googlebatch-mount-path` | str | | False | /mnt/share | |
90 | | -| work_tasks | The default number of work tasks (these are NOT MPI ranks) | `--googlebatch-work-tasks` | int | | False | 1 | |
91 | | -| cpu_milli | Milliseconds per cpu-second | `--googlebatch-cpu-milli` | int | | False | 1000 | |
92 | | -| work_tasks_per_node | The default number of work tasks per node (Google Batch calls these tasks) | `--googlebatch-work-tasks-per-node` | int | | False | 1 | |
93 | | -| memory | Memory in MiB | `--googlebatch-memory` | int | | False | 1000 | |
94 | | -| retry_count | Retry count (default to 1) | `--googlebatch-retry-count` | int | | False | 1 | |
95 | | -| max_run_duration | Maximum run duration, string (e.g., 3600s) | `--googlebatch-max-run-duration` | str | | False | "3600s" | |
96 | | -| labels | Comma separated key value pairs to label job (e.g., model=a3,stage=test) |`--googlebatch-labels` | str | | False | unset| |
97 | | -| container | Container to use (only when image_family is batch-cos*) [see here](https://cloud.google.com/batch/docs/vm-os-environment-overview#supported_vm_os_images) for families/projects | `--googlebatch-container` | str | | False | unset| |
98 | | -| keep_source_cache | Cache workflows in your Google Cloud Storage Bucket | `--googlebatch-keep-source-cache` | bool | | False | False | |
99 | | -| snippet | A comma separated list of one or more snippets to add to your setup | `--googelbatch-snippets` | str | | False | unset | |
100 | | - |
101 | | -For machine type, note that for MPI workloads, mpitune configurations are validated on c2 and c2d instances only. |
102 | | -Also note that you can customize the machine type on the level of the step (see [Step Options](#step-options) below). |
103 | | - |
104 | | -#### Choosing an Image |
105 | | - |
106 | | -You can read about how to choose an image [here](https://cloud.google.com/batch/docs/view-os-images). Note that |
107 | | -the image family and project must match or you'll see that your job does not run (but has an event that indicates a mismatch in the online table). |
108 | | -Since this is a changing set we do not validate, however we suggest that you check before running to not waste time. |
109 | | -I am not entirely sure how to choose correctly, because there is some information [here]() but this listing offers |
110 | | -different information: |
111 | | - |
112 | | -```bash |
113 | | -gcloud compute images list | grep cos |
114 | | -``` |
115 | | - |
116 | | -### Batch Snippets |
117 | | - |
118 | | -Batch, by way of running on virtual machines, can support custom more complex setups or running steps such as running MPI. |
119 | | -However, the setups here are non trivial, so if you choose, a custom snippet can be added. There are |
120 | | -two types of snippets: |
121 | | - |
122 | | - - named, built-in snippets provided by the googlebatch executor plugin here |
123 | | - - your custom snippet provided via a script file (not implemented yet) |
124 | | - |
125 | | -For each named snippet, depending on the functionality it might add custom logic to the setup or final runnable step. |
126 | | -Examples for providing both are shown below. To determine if the snippet is custom, it should be a json or yaml file that |
127 | | -exists. The order that you provide any number of snippets is the order they |
128 | | -are added. To provide more than one, provide them via a comma separated list. |
129 | | - |
130 | | -```bash |
131 | | -$ snakemake --jobs 1 --executor googlebatch --googlebatch-bucket snakemake-cache-dinosaur --googlebatch-snippets intel-mpi |
132 | | -``` |
133 | | - |
134 | | -### Additional Environment Variables |
135 | | - |
136 | | -The following environment variables are available within any Google batch run: |
137 | | - |
138 | | - - `BATCH_TASK_INDEX`: The index of the workflow step (Google Batch calls a "task") |
139 | | - |
140 | | -### Step Options |
141 | | - |
142 | | -The following options are allowed for batch steps. This predominantly includes most arguments. |
143 | | - |
144 | | -#### googlebatch_machine_type |
145 | | - |
146 | | -This will define the machine type for a particular step, overriding the default from the command line. |
147 | | - |
148 | | -```console |
149 | | -rule hello_world: |
150 | | - output: |
151 | | - "...", |
152 | | - resources: |
153 | | - googlebatch_machine_type="c3-standard-112" |
154 | | - shell: |
155 | | - "..." |
156 | | -``` |
157 | | - |
158 | | -#### googlebatch_image_family |
159 | | - |
160 | | -This will define the image family for a particular step, overriding the default from the command line. |
161 | | - |
162 | | -```console |
163 | | -rule hello_world: |
164 | | - output: |
165 | | - "...", |
166 | | - resources: |
167 | | - googlebatch_image_family="hpc-centos-7" |
168 | | - shell: |
169 | | - "..." |
170 | | -``` |
171 | | - |
172 | | - |
173 | | -#### googlebatch_image_project |
174 | | - |
175 | | -This will define the image project for a particular step, overriding the default from the command line. |
176 | | - |
177 | | -```console |
178 | | -rule hello_world: |
179 | | - output: |
180 | | - "...", |
181 | | - resources: |
182 | | - googlebatch_image_project="cloud-hpc-image-public" |
183 | | - shell: |
184 | | - "..." |
185 | | -``` |
186 | | - |
187 | | - |
188 | | -#### googlebatch_bucket |
189 | | - |
190 | | -This will define the bucket for a particular step, overriding the default from the command line. |
191 | | - |
192 | | -```console |
193 | | -rule hello_world: |
194 | | - output: |
195 | | - "...", |
196 | | - resources: |
197 | | - googlebatch_bucket="my-snakemake-batch-bucket" |
198 | | - shell: |
199 | | - "..." |
200 | | -``` |
201 | | - |
202 | | -#### googlebatch_mount_path |
203 | | - |
204 | | -This will define the mount path for a bucket for a particular step, overriding the default from the command line. |
205 | | - |
206 | | -```console |
207 | | -rule hello_world: |
208 | | - output: |
209 | | - "...", |
210 | | - resources: |
211 | | - googlebatch_mount_path="/mnt/workflow" |
212 | | - shell: |
213 | | - "..." |
214 | | -``` |
215 | | - |
216 | | - |
217 | | -#### googlebatch_work_tasks |
218 | | - |
219 | | -This will define the work tasks for a particular step, overriding the default from the command line. |
220 | | - |
221 | | -```console |
222 | | -rule hello_world: |
223 | | - output: |
224 | | - "...", |
225 | | - resources: |
226 | | - googlebatch_work_tasks=1 |
227 | | - shell: |
228 | | - "..." |
229 | | -``` |
230 | | - |
231 | | -#### googlebatch_cpu_milli |
232 | | - |
233 | | -This will define the milliseconds per cpu-second for a particular step, overriding the default from the command line. |
234 | | - |
235 | | -```console |
236 | | -rule hello_world: |
237 | | - output: |
238 | | - "...", |
239 | | - resources: |
240 | | - googlebatch_cpu_mulli=2000 |
241 | | - shell: |
242 | | - "..." |
243 | | -``` |
244 | | - |
245 | | -#### googlebatch_work_tasks_per_node |
246 | | - |
247 | | -This will define the work tasks per node (Google batch calls these tasks) for a particular step, overriding the default from the command line. |
248 | | - |
249 | | -```console |
250 | | -rule hello_world: |
251 | | - output: |
252 | | - "...", |
253 | | - resources: |
254 | | - googlebatch_work_tasks_per_node=2 |
255 | | - shell: |
256 | | - "..." |
257 | | -``` |
258 | | - |
259 | | -#### googlebatch_memory |
260 | | - |
261 | | -This will define the memory for a particular step as an integer in MiB, overriding the default from the command line. |
262 | | - |
263 | | -```console |
264 | | -rule hello_world: |
265 | | - output: |
266 | | - "...", |
267 | | - resources: |
268 | | - googlebatch_memory=2000 |
269 | | - shell: |
270 | | - "..." |
271 | | -``` |
272 | | - |
273 | | - |
274 | | -#### googlebatch_retry_count |
275 | | - |
276 | | -This will define the retry times for a step overriding the default from the command line. |
277 | | - |
278 | | -```console |
279 | | -rule hello_world: |
280 | | - output: |
281 | | - "...", |
282 | | - resources: |
283 | | - googlebatch_retry_count=2 |
284 | | - shell: |
285 | | - "..." |
286 | | -``` |
287 | | - |
288 | | -#### googlebatch_max_run_duration |
289 | | - |
290 | | -This will define the max run duration for a step overriding the default from the command line. |
291 | | - |
292 | | -```console |
293 | | -rule hello_world: |
294 | | - output: |
295 | | - "...", |
296 | | - resources: |
297 | | - googlebatch_max_run_duration="3600s" |
298 | | - shell: |
299 | | - "..." |
300 | | -``` |
301 | | - |
302 | | -#### googlebatch_labels |
303 | | - |
304 | | -This will define the extra labels to add to the Google Batch job. |
305 | | - |
306 | | -```console |
307 | | -rule hello_world: |
308 | | - output: |
309 | | - "...", |
310 | | - resources: |
311 | | - googlebatch_labels="model=c3,stage=test" |
312 | | - shell: |
313 | | - "..." |
314 | | -``` |
315 | | - |
316 | | - |
317 | | -#### googlebatch_container |
318 | | - |
319 | | -A container to use only with `image_family` set to batch-cos* (see [here](https://cloud.google.com/batch/docs/vm-os-environment-overview#supported_vm_os_images) for how to see VM choices) |
320 | | - |
321 | | -```console |
322 | | -rule hello_world: |
323 | | - output: |
324 | | - "...", |
325 | | - resources: |
326 | | - googlebatch_container="ghcr.io/rse-ops/atacseq:app-latest" |
327 | | - shell: |
328 | | - "..." |
329 | | -``` |
330 | | - |
331 | | -#### googlebatch_snippets |
332 | | - |
333 | | -One or more named (or file-derived) snippets to add to setup. |
334 | | - |
335 | | -```console |
336 | | -rule hello_world: |
337 | | - output: |
338 | | - "...", |
339 | | - resources: |
340 | | - googlebatch_snippets="mpi,myscript.sh" |
341 | | - shell: |
342 | | - "..." |
343 | | -``` |
| 4 | +For documentation, see the [Snakemake plugin catalog](https://snakemake.github.io/snakemake-plugin-catalog/plugins/executor/googlebatch.html). |
344 | 5 |
|
345 | 6 | ### TODO |
346 | 7 |
|
|
0 commit comments