[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [tlaplus] TLC: NoSuchElementException on checkpoint restore
On 09.12.2015 14:09, Jaak Ristioja wrote:
> 3 sets, one for each slave.
combining N fingerprint sets of a distributed TLC run into a single one
is logically possible but technically not supported.
> In my distributed setup, the slaves seems to be rather idle, If I get
> access to a more powerful machine I want to try whether doing the
> computation locally on that machine would be any faster. I suspect that
> besides the master CPU usage, networking between the master and slaves
> might also be a bottleneck.
> I recorded some of data on the network interface of the master and took
> a look at it using Wireshark. I saw uncompressed RMI data containing
> quite a lot of TLA model identifier strings being passed. I guess that
> networking could be well optimized.
What kind of network are your nodes connected with?