1.1.0 Not working well under Ubuntu 14.04 (memory leak?)

I updated to 1.1.0 using the apt distribution. While the daemon starts correctly and seems to operate, it grows in size over time. After 18 hours it is up to a 12 gig process size and has ground the machine to a crawl.

Also, it never accepts more than 2 incoming connections. Relatively soon after start up it is up to the usual 8 connections but over the 18 hours that I have tracked it, it never gets more than 2 incoming and usually it has none.

The prior version, on the same machine would run for long periods of time without memory bloating and would hover around 80+ connections.

Which release of Ubuntu e.g 14.04.5 ? Did zcash.conf stay the same when upgrading? What does ‘zcash-cli getmempoolinfo’ show?

Yes the configuration was initially the same after upgrading. Since then I have changed it, trying to figure out what was going on.

It turns out that this memory growth only happens if you turn on mining. I had been mining on two threads. When I turn off mining it seems to be stable.

I have been running it without mining for a day and so far have not seen unusual memory growth. If you need the getmempoolinfo information, I can restart with mining and gather that for you after some period of time.

Release information:

lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 14.04.5 LTS
Release: 14.04
Codename: trusty

With the following configuration:

addnode=mainnet.z.cash
rpcuser=xxxxxxxx
rpcpassword=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
gen=1
genproclimit=2
equihashsolver=tromp
listen=1
maxconnections=100

The process size after 1 hour was 2.5 GB. Now after 15 hours it has grown to 4.8 GB.

$ zcash-cli getmempoolinfo
{
“size”: 27,
“bytes”: 282509,
“usage”: 604016
}

Thanks for the info. We’ll investigate.

@TO99 To clarify, when you called ‘getmempoolinfo’ was the memory usage at 4.8 GB? Or did you call ‘getmempoolinfo’ after a shutdown and restart? I’d like to find out the size of the mempool when memory grows.

The memory was at 4.8GB when I called getmempoolinfo.

That instance is still running. It has now been up 17 hours and something has triggered it to go crazy in the last 2 hours. The process size is now 11.7GB and starting to respond very slowly. I issued a getmempoolinfo command and it has been several minutes and I have not gotten a response yet…

Eventually got a response from getmempoolinfo:

$ zcash-cli getmempoolinfo
{
“size”: 383,
“bytes”: 27919909,
“usage”: 59209744
}