It does seem to work the other way too. That is, I started running Ebola again after several days without it, along with MIP. I got a few errors on Ebola (about one out of ten), but also for the first time I picked up three errors on MIP. I don't recall ever seeing that before when they run by themselves.

One thing I do a little differently than most people is that I run Folding on a GPU (GTX 1070), and reserve a CPU core for it. Normally, there is no interaction with the BOINC projects, but I have seen it on rare occasion. However, I consider it high priority, and will not be stopping it to check. So my experience may not be the same as for others, but yes MIP is a little more problematic than some.

The MIP errors were all the same:

<core_client_version>7.12.0</core_client_version>
<![CDATA[
<message>
process got signal 11</message>
<stderr_txt>
[2018-10-18 20:43:41:] :: BOINC:: Initializing ... ok.
[2018-10-18 20:43:41:] :: BOINC :: boinc_init()
INFO: result number = 0
BOINC:: Setting up shared resources ... ok.
BOINC:: Setting up semaphores ... ok.
BOINC:: Updating status ... ok.
BOINC:: Registering timer callback... ok.
BOINC:: Worker initialized successfully.
command: ../../projects/www.worldcommunitygrid.org/wcgrid_mip1_rosetta_7.16_x86_64-pc-linux-gnu -in::file::zip MIP1_databasev2.zip @./MIP1_00109580.flags -out::file::silent result_silent.out -run:jran 1309508479 -nstruct 3 -out::level 100 -run::no_scorefile true
Registering options..
Registered extra options.
Initializing broker options ...
Registered extra options.
Initializing core...
Initializing options.... ok
Options::initialize()
Options::adding_options()
Options::initialize() Check specs.
Options::initialize() End reached
Loaded options.... ok
Processed options.... ok
Initializing random generators... ok
Initialization complete.
Setting WU description ...
Unpacking zip data: ../../projects/www.worldcommunitygrid.org/mip1.MIP1_databasev2.zip
Setting database description ...
Setting up checkpointing ...
Setting up graphics native ...
set_shared_memory_fully_initialized ...
abrelax ...
abrelax.run
Setting up folding (abrelax) ...
Beginning folding (abrelax) ...
BOINC:: Worker startup.
Sequence Length = 223
Starting work on structure: _0001
Finished _0001 in 1984.02 seconds.
Starting work on structure: _0002

</stderr_txt>

Interesting. I started using a ramdisk or write-cache back in the CEP2 days (to protect the SSD), and it generally avoided problems. I still use a 12 GB write cache on the Ryzen 1700 (Ubuntu 18.04). But with 32 GB main memory, I still have 22 GB free at the moment (not all the cache is used). That is with four MIP running, and all the other cores busy on WCG or Folding/GPU. So it is probably some resource other than memory in conflict, though I have no idea what.

I actually limit MIP to four at a time with an app_config to prevent problems with run times; they are currently averaging 1 hour 30 minutes, and the maximum is under 3 hours. Maybe I will try limiting them to three at a time, though I think I will just drop Ebola first. It is near the end of its run, and they don't really need me now.

Thanks.