Multiple gpg-agent processes

Troubleshooting
1

Both my production Rockstor, and my test Rockstor VM are suffering from multiple instances of gpg-agent and scdaemon proliferating as well as /var/tmp filling up with temporary files.

.....
24885 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.0jQCIQ/zypp-trusted-kr3ypwAS --use-standard-socket --daemon
24887 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.0jQCIQ/zypp-trusted-kr3ypwAS
11246 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.nWpD2p/zypp-trusted-krUcyPpK --use-standard-socket --daemon
11248 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.nWpD2p/zypp-trusted-krUcyPpK
29726 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.J3v1ss/zypp-trusted-krdZNh2h --use-standard-socket --daemon
29728 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.J3v1ss/zypp-trusted-krdZNh2h
16192 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.BObu1D/zypp-trusted-krcM5dgO --use-standard-socket --daemon
16194 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.BObu1D/zypp-trusted-krcM5dgO
 2477 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.vfxOpN/zypp-trusted-krrwxS4U --use-standard-socket --daemon
 2479 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.vfxOpN/zypp-trusted-krrwxS4U
21140 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.INcmQe/zypp-trusted-krloifAs --use-standard-socket --daemon
21142 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.INcmQe/zypp-trusted-krloifAs
 7286 ?        Ssl    0:02 gpg-agent --homedir /var/tmp/zypp.ITAKY0/zypp-trusted-krXK4fK9 --use-standard-socket --daemon
 7288 ?        SLl    0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.ITAKY0/zypp-trusted-krXK4fK9
.......

These processes are all owned by the rockstor service and judging by the file names are to do with checking for zypper updates.

If I leave my NAS running long enough without a reboot everything starts slowing down, in particular my Adguard Rockon gets very slow.

Any suggestions on how to stop them?

1 Like
2

@delboy Hello again, and thanks for another nice report.

I’ve reproduced this issue with a fresh install on our 5.1.0-0 based installer and made reference to my initial impressions of the cause in the following issue:

When one refreshes our dashboard (via the browser), we make a call to zypper that currently often times-out. That seems like a likely candidate. Where a potential work-around is to avoid a large backlog of upstream updates. But there may well be a fix already committed re our internal time-outs. But this has yet to be solidly connected, and is as-yet unreleased in rpm form. But our first rpm in the next testing phase should help with testing this.

@delboy Take a look at that issue and see what you think. Nice find by the way. And thanks again for reporting.

The reproducer system used in that issue now shows only the following:

ps -aux --forest | grep "gpg-agent\|scdaemon"
root     16968  0.0  0.0   8380  2304 pts/0    S+   17:16   0:00              \_ grep gpg-agent\|scdaemon
root      2591  0.0  0.1 245576  5704 ?        Ssl  17:05   0:00 gpg-agent --homedir /root/.gnupg --use-standard-socket --daemon
root      2596  0.0  0.2 109116  8972 ?        SLl  17:05   0:00  \_ scdaemon --multi-server

With a transient additional entry lasting only a few seconds as follows:

root     17156  0.0  0.0 171844  3640 ?        Ssl  17:18   0:00 gpg-agent --homedir /var/tmp/zypp.lmfPcx/zypp-trusted-krRiTbID --use-standard-socket --daemon
root     17158  0.0  0.2 100920  8840 ?        SLl  17:18   0:00  \_ scdaemon --multi-server --homedir /var/tmp/zypp.lmfPcx/zypp-trusted-krRiTbID

showing up on every manual browser refresh of our dashboard.

Hope that helps.

1 Like