Hey,

I was using dmtcp-2.3.1 with OpenMPI-1.8.6 to create checkpoints on
parallel application. It worked fine till restart, when he couldn't find
file in 'tmp' directory. I tried the same with dmtcp-2.4.0-rc5 and there it
works great.

However option --enable-unique-checkpoint-filenames doesn't work. I get the
following response:

NOTE at dmtcp_coordinator.cpp:1291 in startCheckpoint; REASON='starting
checkpoint, suspending all nodes'
     s.numPeers = 1
[3592] NOTE at dmtcp_coordinator.cpp:1293 in startCheckpoint;
REASON='Incremented Generation'
     compId.generation() = 11
[3592] NOTE at dmtcp_coordinator.cpp:654 in updateMinimumState;
REASON='locking all nodes'
[3592] NOTE at dmtcp_coordinator.cpp:660 in updateMinimumState;
REASON='draining all nodes'
[3592] NOTE at dmtcp_coordinator.cpp:666 in updateMinimumState;
REASON='checkpointing all nodes'
[3592] WARNING at dmtcp_coordinator.cpp:1562 in writeRestartScript;
REASON='JWARNING(symlinkat(uniqueFilename.c_str(), dirfd, filename.c_str())
== 0) failed'
[3592] NOTE at dmtcp_coordinator.cpp:680 in updateMinimumState;
REASON='building name service database'
[3592] NOTE at dmtcp_coordinator.cpp:696 in updateMinimumState;
REASON='entertaining queries now'
[3592] NOTE at dmtcp_coordinator.cpp:701 in updateMinimumState;
REASON='refilling all nodes'
[3592] NOTE at dmtcp_coordinator.cpp:732 in updateMinimumState;
REASON='restarting all nodes'
JTIMER(checkpoint) : 1.642

The thing is that even with serial application i get this warning and at
last i have only the newest checkpoint file.

Thanks in advance for your help!
Michal
------------------------------------------------------------------------------
Don't Limit Your Business. Reach for the Cloud.
GigeNET's Cloud Solutions provide you with the tools and support that
you need to offload your IT needs and focus on growing your business.
Configured For All Businesses. Start Your Cloud Today.
https://www.gigenetcloud.com/
_______________________________________________
Dmtcp-forum mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dmtcp-forum

Reply via email to