「RELIONトラブルシューティング」の版間の差分

提供: Eospedia
移動: 案内検索
行27: 行27:
  
 
(普通は上記だけでlibtiffをリンクしたRELIONをビルドできるはずです。ただ、minicondaやanacondaを使ってEMAN2をビルドした環境とかでは、[[minicondaにCMakeの検索パスを持ってかれる]]というつらい現象が起きるかもしれません。)
 
(普通は上記だけでlibtiffをリンクしたRELIONをビルドできるはずです。ただ、minicondaやanacondaを使ってEMAN2をビルドした環境とかでは、[[minicondaにCMakeの検索パスを持ってかれる]]というつらい現象が起きるかもしれません。)
 +
 +
 +
== 2D classification ==
 +
===<u>[[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129</u>===
 +
==== 症状 ====
 +
2019/5/21, v3.0.5, build from source in Ubuntu 16.04.6 LTS
 +
 +
RELION3チュートリアルで最初の2D classificationを実行したらMPI関連のエラーが出て異常終了した。
 +
 +
エラー全文は以下。DL-Boxはホストマシン名。
 +
<pre>
 +
[DL-Box:00925] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
 +
[DL-Box:00926] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
 +
[DL-Box:00927] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
 +
[DL-Box:00928] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
 +
*** An error occurred in MPI_Init
 +
*** on a NULL communicator
 +
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
 +
***    and potentially your MPI job)
 +
*** An error occurred in MPI_Init
 +
*** on a NULL communicator
 +
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
 +
***    and potentially your MPI job)
 +
*** An error occurred in MPI_Init
 +
*** on a NULL communicator
 +
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
 +
***    and potentially your MPI job)
 +
*** An error occurred in MPI_Init
 +
*** on a NULL communicator
 +
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
 +
***    and potentially your MPI job)
 +
[DL-Box:926] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
 +
[DL-Box:925] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
 +
[DL-Box:928] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
 +
[DL-Box:927] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
 +
[DL-Box:00929] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
 +
*** An error occurred in MPI_Init
 +
*** on a NULL communicator
 +
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
 +
***    and potentially your MPI job)
 +
[DL-Box:929] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
 +
</pre>
 +
 +
 +
ジョブのパラメータのうちMPIに関係するかもしれない部分は以下(I/Oとかも含め)
 +
<pre>
 +
Combine iterations through disc? == No
 +
Use parallel disc I/O? == Yes
 +
Pre-read all particles into RAM? == Yes
 +
Which GPUs to use: == 0:1:2:3
 +
Minimum dedicated cores per node: == 1
 +
Number of MPI procs: == 5
 +
Number of pooled particles: == 30
 +
Number of threads: == 3
 +
Copy particles to scratch directory: ==
 +
Use GPU acceleration? == Yes
 +
</pre>

2019年5月21日 (火) 08:36時点における版

Motion Correction

ERROR: TIFF support was not enabled during compilation

症状

2019/5/14, v3.0.5, build from source in Ubuntu 16.04.6 LTS

RELIONチュートリアルでMotion correctionするとき、MoitonCor2のRELION実装使ったら表題のエラーが出た。

対処

# relionのルートディレクトリに移動 (cmakeとかsrcとかあるディレクトリ)

$ rm -r build/ install/

$ sudo apt install libtiff5-dev

$ mkdir build/ install/

$ cd build && cmake -DCMAKE_INSTALL_PREFIX=../install ..

$ make -j10

$ make install

(普通は上記だけでlibtiffをリンクしたRELIONをビルドできるはずです。ただ、minicondaやanacondaを使ってEMAN2をビルドした環境とかでは、minicondaにCMakeの検索パスを持ってかれるというつらい現象が起きるかもしれません。)


2D classification

[[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129

症状

2019/5/21, v3.0.5, build from source in Ubuntu 16.04.6 LTS

RELION3チュートリアルで最初の2D classificationを実行したらMPI関連のエラーが出て異常終了した。

エラー全文は以下。DL-Boxはホストマシン名。

[DL-Box:00925] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
[DL-Box:00926] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
[DL-Box:00927] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
[DL-Box:00928] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
*** An error occurred in MPI_Init
*** on a NULL communicator
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
***    and potentially your MPI job)
*** An error occurred in MPI_Init
*** on a NULL communicator
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
***    and potentially your MPI job)
*** An error occurred in MPI_Init
*** on a NULL communicator
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
***    and potentially your MPI job)
*** An error occurred in MPI_Init
*** on a NULL communicator
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
***    and potentially your MPI job)
[DL-Box:926] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
[DL-Box:925] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
[DL-Box:928] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
[DL-Box:927] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!
[DL-Box:00929] [[INVALID],INVALID] ORTE_ERROR_LOG: Not found in file runtime/orte_init.c at line 129
*** An error occurred in MPI_Init
*** on a NULL communicator
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
***    and potentially your MPI job)
[DL-Box:929] Local abort before MPI_INIT completed successfully; not able to aggregate error messages, and not able to guarantee that all other processes were killed!


ジョブのパラメータのうちMPIに関係するかもしれない部分は以下(I/Oとかも含め)

Combine iterations through disc? == No
Use parallel disc I/O? == Yes
Pre-read all particles into RAM? == Yes
Which GPUs to use: == 0:1:2:3
Minimum dedicated cores per node: == 1
Number of MPI procs: == 5
Number of pooled particles: == 30
Number of threads: == 3
Copy particles to scratch directory: ==
Use GPU acceleration? == Yes