rvernica / docker-library Goto Github PK

Problem description

Hi, @rvernica.
Many thanks for being so generous with such a useful docker image. I had a problem with rvernica/scidb:19.11-xenial about open mpi.
I ran the docker image to create the container according to the README, and I ran the official example of gemm in the container. However, it produces some errors as shown below.

Console

docker run --tty --interactive rvernica/scidb:19.11-xenial bash

• Starting OpenBSD Secure Shell server sshd [ OK ]
• Starting PostgreSQL 9.3 database server [ OK ]
  shim: SciDB HTTP service started on port(s) 8080,8083s with web root [/var/lib/shim/wwwroot], talking to SciDB on port 1239
  [scidbctl] Starting SciDB cluster scidb ...
  [scidbctl-0-0-scidb] Starting s0-i0 on server 127.0.0.1
  [scidbctl-0-1-scidb] Starting s0-i1 on server 127.0.0.1
  [scidbctl] Started 2 instances, waiting up to 30 seconds for cluster sync
  [scidbctl] Cluster is ready
  root@cb5274826c01:/# iquery --afl
AFL% list('instances');
  {No} name,port,instance_id,online_since,instance_path
  {0} '127.0.0.1',1239,0,'2020-08-05 14:34:37','/opt/scidb/19.11/DB-scidb/0/0'
  {1} '127.0.0.1',1240,1,'2020-08-05 14:34:37','/opt/scidb/19.11/DB-scidb/0/1'
  AFL% load_library('dense_linear_algebra');
  Query was executed successfully
  AFL% CREATE ARRAY A <val:double>[i=0:1; j=0:1];
  Query was executed successfully
  AFL% store(build (A, i*2 + j + 1), A);
  Query was executed successfully
  AFL% store(build(<val:double>[i=0:1; j=0:1],iif(i=j,1,0)),B);
  Query was executed successfully
  AFL% CREATE ARRAY C <val:double>[i=0:1; j=0:1];
  Query was executed successfully
  AFL% store(build (C, 1), C);
  Query was executed successfully
  AFL% gemm(A,B,C);
  SystemException in file: src/mpi/MPISlaveProxy.cpp function: checkLauncher line: 56 instance: s0-i0 (0)
  Error id: scidb::SCIDB_SE_INTERNAL::SCIDB_LE_OPERATION_FAILED
  Error description: Internal SciDB error. Operation 'MPI launcher process already terminated' failed.

config.ini

[scidb]
base-path=/opt/scidb/19.11/DB-scidb
base-port=1239
db_user=scidb
install_root=/opt/scidb/19.11
logconf=/opt/scidb/19.11/share/scidb/log4cxx.properties
pluginsdir= /opt/scidb/19.11/lib/scidb/plugins
server-0=127.0.0.1,1
execution-threads=4
result-prefetch-threads=3
result-prefetch-queue-size=1
operator-threads=1
sg-send-queue-size=4
sg-receive-queue-size=4

scidb.log (server-0, 0, i.e., the first scidb instance)

2022-05-16 14:27:41.000054 [0x7f1bd8393e80] [WARN ]: Ignoring io-paths-list config.ini entry "" (""): Not an absolute path
2022-05-16 14:27:45.000633 [0x7fd1ddcd4e80] [WARN ]: Ignoring io-paths-list config.ini entry "" (""): Not an absolute path
2022-05-16 14:27:45.000651 [0x7fd1ddcd4e80] [ERROR]: Network error #system:111 (Connection refused) while connecting to s0-i1 (1), 127.0.0.1:1240
2022-05-16 14:27:45.000651 [0x7fd1ddcd4e80] [ERROR]: NetworkManager::handleConnectionError: Connection error - aborting ALL queries
2022-05-16 14:33:07.000541 [0x7fd1cd367700] [ERROR]: completeExecuteQuery failed to complete for queryID=0.1652711583535385739 : SystemException in file: src/mpi/MPISlaveProxy.cpp function: checkLauncher line: 56 instance: s0-i0 (0)
Error id: scidb::SCIDB_SE_INTERNAL::SCIDB_LE_OPERATION_FAILED
Error description: Internal SciDB error. Operation 'MPI launcher process already terminated' failed.
2022-05-16 14:33:07.000541 [0x7fd1cd367700] [ERROR]: Query (0.1652711583535385739) error handlers (2) are being executed
2022-05-16 14:33:07.000542 [0x7fd1cd367700] [WARN ]: MpiErrorHandler::cleanupSlavePidFile: cannot read pids for: /opt/scidb/19.11/DB-scidb/0/0/mpi_pid/0.1652711583535385739.1.mpi_slave_scidb
2022-05-16 14:33:07.000542 [0x7fd1cd367700] [ERROR]: SciDB MPI launcher (pid=-1704) exited with status = 1
2022-05-16 14:33:07.000542 [0x7fd1cd367700] [ERROR]: Failed to destroy launcher for launch = 1 (0.1652711583535385739) because: SystemException in file: src/mpi/MPILauncher.cpp function: completeLaunch line: 283 instance: s0-i0 (0)
Error id: scidb::SCIDB_SE_INTERNAL::SCIDB_LE_OPERATION_FAILED
Error description: Internal SciDB error. Operation 'MPI launcher process failed' failed.

Methods I have tried

This problem has been troubling me for two days. I searched for many solutions but none of them worked.
(1) ssh, follow the advice from the official manual.

(2) firewall, ufw allow 1240.

(3) telnet 127.0.0.1 1240

(4) the enough shared memory available, follow the advice from the official manual, df -h /dev/shm --> 90 GB.

Hi, thanks for creating this docker container for SciDB. I was trying to use it but ran into an issue mainly that Postgres can't be setup because a Permission check fails on /root/.pgpass. This is the output I was getting after running docker run --tty rvernica/scidb:16.9 bash

[ ok ] Starting OpenBSD Secure Shell server: sshd.
[ ok ] Starting PostgreSQL 9.4 database server: main.
Starting shim
shim: SciDB HTTP service started on port(s) 8080,8083s with web root [/var/lib/shim/wwwroot], talking to SciDB on port 1239
scidb.py: WARNING: PgpassUpdater: Permission check failed on /root/.pgpass
Postgres user: scidb
Postgres password [scidb]: 

If it helps I can provide any additional details you may need.

@rvernica

Putting in a request for scidb:19.11-ext Docker (similar to my request #10 for scidb:19.3-ext)

FYI -- extra-scidb-libs for 19.11 is now available at this branch (docs might need to be updated).

The following was tested after the following setup:

• Run scidb 19.11 ext centos 7 docker image to get scidb 19.11.3 CE
• run yum update extra-scidb-libs-19.11 to get extra-scidb-libs 19.11-2

Problem description

On scidb 19.11.3 CE and extra-scidb-libs 19.11-2, the following R script hangs randomly.

# in R
library(scidb)
db = scidbconnect()

N = 100  # try increasing this if it does not hang within 100 tries
for (idx in 1:N){
  ti = as.numeric(system.time({
    iquery(db, "create array temp <val:string>[i=0:0]")
    iquery(db, "remove(temp)")
  })['elapsed'])
  message("Attempt: ", idx, " -- ", format(round(ti, 2), nsmall = 2), " seconds")
}

Version info

scidb --version
# SciDB Version: 19.11.3
# Build Type: RelWithDebInfo
# Commit: e75cd16
# Copyright (C) 2008-2019 SciDB, Inc.

yum list installed | grep extra-scidb-libs
# extra-scidb-libs-19.11.x86_64             2-1                    @scidb-extra

cat /etc/centos-release
# CentOS Linux release 7.7.1908 (Core)

@rvernica

I depend on scidb:18.1-ext docker library for my revealgenomics Travis framework. How hard would it be to build the same for 19.3? I saw that you already have a few 19.3 Docker images, but not the -ext one.

Apply parallel_binary_load patch provided by Paradigm4:

https://drive.google.com/folderview?id=0B8eyzr2ndWOTUmRsZUJoQU1tTmc&usp=drive_web&tid=0B7yt0n33Us0rT1FJdmxFV2g0OHc#list