System

Cluster state and node information

serve mode · sqlite · 127.0.0.1:8080
Version
1.1.0
alpha.1 · linux/amd64
Uptime
3d 14h
since Jul 30, 09:12
Active jobs
2
1 running · 1 waiting
Workers online
3 / 5
2 busy · 1 idle

Storage usage

2.1 GB total
Datasets · 812 MB Artifacts · 1.2 GB Database · 34 MB

Health

all checks pass
Database
Connected sqlite · wal
Userservice
Embedded 127.0.0.1:41273
Reducer
Idle last merge 12 min ago
Node information
Binary
/usr/local/bin/coordinator
Listen address
127.0.0.1:8080
Data directory
~/.scimesh (jwt.secret · worker.token · venv)
Database engine
sqlite · ~/.scimesh/scimesh.db
Public URL
http://192.168.1.10:8080
JobWorkloadOwnerStatusProgressSubmitted
kinase-inhibitors-topk
7f3a…c91e
similarity-searchalice@lab.orgRunning
124 / 200 shards
12 min ago
chembl-2d-descriptors
a102…77bd
descriptorsbob@lab.orgWaiting
0 / 48 shards
4 min ago
fragment-library-graph
55e0…0a3f
similarity-graphalice@lab.orgCompleted
96 / 96 shards · 3m 41s
1 h ago
leadopt-molwt-450
9b41…e2c8
molwt-filteradmin@scimesh.localCompleted
12 / 12 shards · 24s
3 h ago
hts-plate-14-search
c7d9…4b60
similarity-searchbob@lab.orgFailed
18 / 53 shards · 2 failed
yesterday
Workers register themselves. Trust controls whether a machine may claim tasks from this cluster.
WorkerStatusCapabilitiesTrustCompletedLast signal
lab-node-01
w-01HZZ…a4
Busysimilarity-searchsimilarity-graph3414 s ago
emil-laptop
w-01HZZ…9c
Idlesimilarity-searchdescriptorsmolwt-filter872 s ago
lab-node-02
w-01HZZ…d1
Busysimilarity-search126 s ago
old-workstation
w-01HZX…7f
Offlinesimilarity-search1 2042 days ago
Users
EmailRoleJobsCreatedLast active
admin@scimesh.local
admin9Jul 30now
alice@lab.org
11Jul 3012 min ago
bob@lab.org
6Jul 311 h ago
Worker keys
Keys let lab machines register as workers under a user account. Served instances can also use the cluster token.
LabelKey prefixOwnerCreatedLast used
lab-node-01smk_7f3a…alice@lab.orgJul 304 s ago
emil-laptopsmk_b291…admin@scimesh.localJul 312 s ago
old-workstationsmk_44cd…alice@lab.orgJul 302 days agoRevoked
Disabled workloads are rejected at submit time and hidden from the job form. The catalog is compiled into the binary; settings persist in the database.
WorkloadReductionParametersDataset uploadEnabled
similarity-search
top-k Tanimoto search over a TSV of SMILES
top-k4 declaredready
similarity-graph
threshold graph edges between all pairs
ordered-concat2 declaredready
descriptors
2D RDKit descriptors per molecule
ordered-concat1 declaredready
molwt-filter
keep molecules inside a weight window
ordered-concat2 declaredready
Jobs · 7 days
26
+8 vs previous week
Shards completed
1 644
across 5 workers
Avg shard time
2.3s
−0.4s this week
Failure rate
1.2%
20 failed shards

Jobs per day

last 7 days
MonTueWed ThuFriSat Sun

Jobs by workload

all time
similarity-search
similarity-graph
molwt-filter
descriptors
The cluster token below authenticates any worker. Reveal it only on a trusted machine, and rotate it if it may have leaked.
Cluster
Worker token
••••••••••••••••••••••••
Public URL
Listen address
takes effect after restart
Storage
Data directory
~/.scimesh
Database engine
Auto-migrate on start
Danger zone
Prune artifacts
Delete results of completed jobs older than 30 days · frees ~400 MB
Reset cluster
Drop all jobs, workers and keys. Users are kept.