issues
search
conveyal
/
otpa-cluster
Cluster backend for otpa many-to-many queries.
0
stars
1
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Cancel multipoint jobs
#47
mattwigway
opened
9 years ago
0
Single point graphs can be booted out
#46
mattwigway
opened
9 years ago
0
Workers can spin
#45
mattwigway
opened
9 years ago
0
thin workers
#44
mattwigway
opened
9 years ago
15
pre-evict graph
#43
mattwigway
opened
9 years ago
1
Lower default chunk size
#42
mattwigway
opened
9 years ago
0
Don't let workers get stuck when graph build fails
#41
mattwigway
opened
9 years ago
0
Duplicated workresults
#40
mattwigway
closed
9 years ago
0
compress pointsets in S3 for transfer performance
#39
mattwigway
closed
9 years ago
0
Cache two graphs on workers?
#38
mattwigway
opened
9 years ago
1
Don't crash SPTWorker when a single routing request fails
#37
mattwigway
closed
9 years ago
0
Don't let queue size grow when processing single point requests
#36
mattwigway
closed
9 years ago
0
Don't evict single point graphs that have been used recently
#35
mattwigway
closed
9 years ago
0
Is EarliestArrivalSearch the best search strategy for non-transit modes?
#34
mattwigway
closed
9 years ago
2
Only upload graphs and pointsets to S3 if they are not there already
#33
mattwigway
closed
9 years ago
2
Error kernel sampleset, pointset and graph caching
#32
mattwigway
closed
9 years ago
2
Don't preserve internal state
#31
mattwigway
closed
9 years ago
1
Be resilient to initial association issues
#30
mattwigway
opened
9 years ago
1
Don't use stable paths to identify actors
#29
mattwigway
closed
9 years ago
0
Don't let progress bar go backwards
#28
mattwigway
closed
9 years ago
1
Don't cache all work results in executive
#27
mattwigway
closed
9 years ago
1
Handle worker manager termination
#26
mattwigway
closed
9 years ago
0
Embed graph ID in AnalystClusterRequest so we can always be sure we're processing requests against the correct graph
#25
mattwigway
closed
9 years ago
1
Handle message loss going to workermanager
#24
mattwigway
closed
9 years ago
1
Don't group large clusters of requests together
#23
mattwigway
closed
9 years ago
1
Fix PointSetDatastore to use caching
#22
mattwigway
closed
9 years ago
1
Set chunk size based on moving average of response times
#21
mattwigway
closed
9 years ago
1
Switch to polling for job assignment
#20
mattwigway
closed
9 years ago
0
split jobs return status from only one worker (show 1/n total for items complete/procesed)
#19
kpwebb
closed
9 years ago
2
Document command line options
#18
mattwigway
opened
9 years ago
0
Allow specifying number of threads for worker on command line
#17
mattwigway
closed
9 years ago
1
Request that a graph be sent to a worker without sending a job
#16
mattwigway
opened
9 years ago
0
Single point queries
#15
mattwigway
closed
9 years ago
2
Assign jobs to workers that have the relevant graph loaded
#14
mattwigway
closed
9 years ago
4
cache graphs
#13
mattwigway
closed
9 years ago
4
Spin up spot instances on demand
#12
mattwigway
opened
9 years ago
0
Don't hard code bucket name
#11
mattwigway
closed
9 years ago
0
Depend on an OTP release
#10
mattwigway
opened
9 years ago
1
Keep graphs in memory
#9
mattwigway
closed
9 years ago
1
Support profile routing
#8
mattwigway
closed
9 years ago
0
respond correctly to subscribing a bogus worker
#7
bmander
opened
10 years ago
0
disconnect, reconnect to live view of running job
#6
bmander
closed
9 years ago
1
query master for result in tidy package
#5
bmander
closed
9 years ago
0
web query the master for job status list
#4
bmander
closed
9 years ago
0
web query the master for worker status list
#3
bmander
closed
9 years ago
0
master disconnects without leaving a mess in the worker
#2
bmander
closed
9 years ago
1
worker disconnects without leaving a mess in the master
#1
bmander
closed
10 years ago
1