1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
|
use std::path::PathBuf;
use std::result::Result as RResult;
use std::sync::Arc;
use anyhow::Context;
use anyhow::Error;
use anyhow::Result;
use anyhow::anyhow;
use diesel::PgConnection;
use futures::FutureExt;
use indicatif::ProgressBar;
use itertools::Itertools;
use tokio::stream::StreamExt;
use tokio::io::AsyncWriteExt;
use tokio::sync::RwLock;
use tokio::sync::mpsc::UnboundedReceiver;
use uuid::Uuid;
use crate::db::models as dbmodels;
use crate::endpoint::Endpoint;
use crate::endpoint::EndpointConfiguration;
use crate::filestore::StagingStore;
use crate::job::JobResource;
use crate::job::RunnableJob;
use crate::log::LogItem;
use crate::util::progress::ProgressBars;
use crate::endpoint::ContainerError;
pub struct EndpointScheduler {
log_dir: Option<PathBuf>,
endpoints: Vec<Arc<RwLock<Endpoint>>>,
staging_store: Arc<RwLock<StagingStore>>,
db: Arc<PgConnection>,
progressbars: ProgressBars,
submit: crate::db::models::Submit,
}
impl EndpointScheduler {
pub async fn setup(endpoints: Vec<EndpointConfiguration>, staging_store: Arc<RwLock<StagingStore>>, db: Arc<PgConnection>, progressbars: ProgressBars, submit: crate::db::models::Submit, log_dir: Option<PathBuf>) -> Result<Self> {
let endpoints = Self::setup_endpoints(endpoints).await?;
Ok(EndpointScheduler {
log_dir,
endpoints,
staging_store,
db,
progressbars,
submit,
})
}
async fn setup_endpoints(endpoints: Vec<EndpointConfiguration>) -> Result<Vec<Arc<RwLock<Endpoint>>>> {
let unordered = futures::stream::FuturesUnordered::new();
for cfg in endpoints.into_iter() {
unordered.push({
Endpoint::setup(cfg)
.map(|r_ep| {
r_ep.map(RwLock::new)
.map(Arc::new)
})
});
}
unordered.collect().await
}
/// Schedule a Job
///
/// # Warning
///
/// This function blocks as long as there is no free endpoint available!
pub async fn schedule_job(&self, job: RunnableJob, multibar: Arc<indicatif::MultiProgress>) -> Result<JobHandle> {
let endpoint = self.select_free_endpoint().await?;
Ok(JobHandle {
log_dir: self.log_dir.clone(),
bar: multibar.add(self.progressbars.job_bar(job.uuid())),
endpoint,
job,
staging_store: self.staging_store.clone(),
db: self.db.clone(),
submit: self.submit.clone(),
})
}
async fn select_free_endpoint(&self) -> Result<Arc<RwLock<Endpoint>>> {
loop {
let ep = self.endpoints
.iter()
.cloned()
.map(|ep| async move {
ep.write()
.await
.number_of_running_containers()
.await
.map(|num_running| (num_running, ep.clone()))
})
.collect::<futures::stream::FuturesUnordered<_>>()
.collect::<Result<Vec<_>>>()
.await?
.iter()
.sorted_by(|tpla, tplb| tpla.0.cmp(&tplb.0))
.map(|tpl| tpl.1.clone())
.next();
if let Some(endpoint) = ep {
return Ok(endpoint)
} else {
trace!("No free endpoint found, retry...");
}
}
}
}
pub struct JobHandle {
log_dir: Option<PathBuf>,
endpoint: Arc<RwLock<Endpoint>>,
job: RunnableJob,
bar: ProgressBar,
db: Arc<PgConnection>,
staging_store: Arc<RwLock<StagingStore>>,
submit: crate::db::models::Submit,
}
impl std::fmt::Debug for JobHandle {
fn fmt(&self, f: &mut std::fmt::Formatter) -> std::result::Result<(), std::fmt::Error> {
write!(f, "JobHandle ( job: {} )", self.job.uuid())
}
}
impl JobHandle {
pub async fn run(self) -> RResult<Vec<dbmodels::Artifact>, ContainerError> {
use crate::db::models as dbmodels;
let (log_sender, log_receiver) = tokio::sync::mpsc::unbounded_channel::<LogItem>();
let ep = self.endpoint.read().await;
let endpoint = dbmodels::Endpoint::create_or_fetch(&self.db, ep.name())?;
let package = dbmodels::Package::create_or_fetch(&self.db, self.job.package())?;
let image = dbmodels::Image::create_or_fetch(&self.db, self.job.image())?;
let envs = self.create_env_in_db()?;
let job_id = self.job.uuid().clone();
trace!("Running on Job {} on Endpoint {}", job_id, ep.name());
let res = ep
.run_job(self.job, log_sender, self.staging_store);
let logres = LogReceiver {
log_dir: self.log_dir.as_ref(),
job_id,
log_receiver,
bar: &self.bar,
}.join();
let (res, logres) = tokio::join!(res, logres);
trace!("Found result for job {}: {:?}", job_id, res);
let log = logres.with_context(|| anyhow!("Collecting logs for job on '{}'", ep.name()))?;
let (paths, container_hash, script) = res.with_context(|| anyhow!("Running job on '{}'", ep.name()))?;
let job = dbmodels::Job::create(&self.db, &job_id, &self.submit, &endpoint, &package, &image, &container_hash, &script, &log)?;
for env in envs {
let _ = dbmodels::JobEnv::create(&self.db, &job, &env)?;
}
// Have to do it the ugly way here because of borrowing semantics
let mut r = vec![];
for p in paths.iter() {
r.push(dbmodels::Artifact::create(&self.db, p, false, &job)?);
}
Ok(r)
}
fn create_env_in_db(&self) -> Result<Vec<dbmodels::EnvVar>> {
trace!("Creating environment in database");
trace!("Hardcoded = {:?}", self.job.package().environment());
trace!("Dynamic = {:?}", self.job.resources());
self.job
.package()
.environment()
.as_ref()
.map(|hm| {
hm.iter()
.inspect(|(k, v)| trace!("Creating environment variable in database: {} = {}", k, v))
.map(|(k, v)| dbmodels::EnvVar::create_or_fetch(&self.db, k, v))
.collect::<Result<Vec<_>>>()
})
.transpose()?
.unwrap_or_default()
.into_iter()
.map(Ok)
.chain({
self.job
.resources()
.iter()
.filter_map(JobResource::env)
.inspect(|(k, v)| trace!("Creating environment variable in database: {} = {}", k, v))
.map(|(k, v)| dbmodels::EnvVar::create_or_fetch(&self.db, k, v))
})
.collect()
}
}
struct LogReceiver<'a> {
log_dir: Option<&'a PathBuf>,
job_id: Uuid,
log_receiver: UnboundedReceiver<LogItem>,
bar: &'a ProgressBar,
}
impl<'a> LogReceiver<'a> {
async fn join(mut self) -> Result<String> {
use resiter::Map;
let mut success = None;
let mut accu = vec!
|