masternode crashes (dumpaddr thread stop)

[doopalek]

Some of my MNs dashd processes sometimes go down for no reason and get delisted even though there is a cron script checking every 30s and starting dashd process if it exits. I really have not idea what is wrong. Please help me. It happens almost every day.

Today 2 masternodes crashed at the same time. Here is log (both have similar logs). Now "masternode debug" command returns: "Not capable masternode: Hot node, waiting for remote activation."
Is there any way to avoid MN to be delisted in this case? I use dash current version, 0.12.0.55

2015-10-01 14:18:25 dstx: Got Masternode transaction xxx
2015-10-01 14:18:25 PrioritiseTransaction: xxx priority += 1000.000000, fee += 0.10
2015-10-01 14:18:48 CActiveMasternode::SendMasternodePing() - Relay Masternode Ping vin = CTxIn(COutPoint(xxx, 1), scriptSig=)
2015-10-01 14:20:24 receive version message: /Dash Core:0.12.0.53/Dash Ninja Port Checker:2.1.1.4/: version 70103, blocks=0, us=[::xxx]:42407, peer=xxx
2015-10-01 14:23:41 receive version message: /Dash Core:0.12.0.55/: version 70103, blocks=116880, us=xxx:9999, peer=xxx
2015-10-01 14:23:42 receive version message: /darkcoinseeder:0.1.2/: version 70103, blocks=320000, us=xxx:9999, peer=xxx
2015-10-01 14:23:53 CActiveMasternode::SendMasternodePing() - Relay Masternode Ping vin = CTxIn(COutPoint(xxx, 1), scriptSig=)
2015-10-01 14:25:39 dumpaddr thread stop
2015-10-01 14:25:39 addcon thread interrupt
2015-10-01 14:25:39 opencon thread interrupt
2015-10-01 14:25:39 msghand thread interrupt
2015-10-01 14:25:39 net thread interrupt
2015-10-01 14:25:39 PrepareShutdown: In progress...
2015-10-01 14:25:39 StopNode()
2015-10-01 14:25:39 Verifying mncache.dat format...
2015-10-01 14:25:41 Loaded info from mncache.dat  2217ms
2015-10-01 14:25:41   Masternodes: 3259, peers who asked us for Masternode list: 1, peers we asked for Masternode list: 8, entries in Masternode list we asked for: 25, nDsqCount: 5042
2015-10-01 14:25:41 Writting info to mncache.dat...
2015-10-01 14:25:42 Written info to mncache.dat  547ms
2015-10-01 14:25:42   Masternodes: 3254, peers who asked us for Masternode list: 0, peers we asked for Masternode list: 0, entries in Masternode list we asked for: 0, nDsqCount: 8495
2015-10-01 14:25:42 Masternode dump finished  2782ms
2015-10-01 14:25:42 Verifying budget.dat format...
2015-10-01 14:25:43 Loaded info from budget.dat  868ms
2015-10-01 14:25:43   Proposals: 9, Budgets: 0, Seen Budgets: 9, Seen Budget Votes: 9208, Seen Final Budgets: 0, Seen Final Budget Votes: 0
2015-10-01 14:25:43 Writting info to budget.dat...
2015-10-01 14:25:43 Written info to budget.dat  154ms
2015-10-01 14:25:43 Budget dump finished  1032ms
2015-10-01 14:25:43 Verifying mnpayments.dat format...
2015-10-01 14:25:43 Loaded info from mnpayments.dat  522ms
2015-10-01 14:25:43   Votes: 40072, Blocks: 4084
2015-10-01 14:25:43 Writting info to mnpayments.dat...
2015-10-01 14:25:44 Written info to mnpayments.dat  156ms
2015-10-01 14:25:44 Budget dump finished  685ms
2015-10-01 14:25:44 Shutdown: done

 

[UdjinM6]

Check "/var/log/syslog" for "dashd" lines. Most likely process was killed by OS because of using too much memory - give it more (increase swap for example).

 

[doopalek]

There is nothing about dashd in syslog and dmsg. :sad: The VPS usually has 300MB free RAM and >2GB free swap... Can I provide anything else to help debug?
Another part of my question is why MN is delisted even though it is restarted by cron?

 

[doopalek]

Increasing swappiness seems to solve this problem.