Connecting worker to master_after multiple attempts

Install, Configure and Update
1

Hello,

I was wondering if anyone could assist. I am working on a standalone station where CryoSparc was installed. When I queued up import movies, I received the following error:

Screenshot_20251210_095839-1
Screenshot_20251210_095839-1727×320 42.2 KB

I attempted multiple times to connect the worker to the master. I also have verified that the ports 39000-39010 are open as well as insured CryoSparc was running. Additionally I have verified the host name using hostname -f and have looked at past discussions on this forum to follow steps taken in the past. Nothing to this point has worked. The error I keep receiving in terminal is below.

I would appreciate any assistance or suggestions one can provide.

Screenshot_20251210_095433
Screenshot_20251210_0954331920×1370 270 KB

Thank you

2

Welcome to the forum @Heather1 .
Please can you post the outputs of these commands (after ensuring CryoSPARC is still running).

host aurora
curl 127.0.0.1:39002
curl aurora:39002
/var/home/localhost/cryosparc/cryosparc_worker/bin/cryosparcw call env | grep -i proxy
3

Good morning @wtempel,

Thank you for your welcome and response.

I have ensured that CryoSparc is still running and attached are the outputs of the commands you provided. Thank you for your help.

Best,

Heather1

Screenshot_20251215_094726
Screenshot_20251215_0947261562×761 81.2 KB

4

@Heather1 Thanks for trying these commands. Under some circumstances, the message CryoSPARC is already running. may be shown even when CryoSPARC processes have been disrupted. To confirm that CryoSPARC processes are indeed running, please can you run some additional commands before retrying the the curl commands

ps -eo user:12,pid,ppid,start,cmd | grep -e cryosparc_ -e mongo
ls -l /tmp/cryosparc*.sock /tmp/mongo*.sock
curl 127.0.0.1:39002
curl aurora:39002
5

@wtempel Thank you for all of your help. I executed the most recent set of commands you provided. Please see the below of the outputs. Again thank you

Screenshot_20251215_125024
Screenshot_20251215_1250241920×504 136 KB

6

@Heather1 These outputs indicate processes and files dated Nov 11 or Dec 8, and suggest that the /tmp/cryosparc-supervisor-*.sock file that should have also been created on Nov 11 may have been (incorrectly) deleted, leading to an inconsistent state of CryoSPARC processes.
You may want to

  1. ensure that /tmp/cryosparc-supervisor-*.sock file is not subject to “improper” deletion or interference. For example, the file may have been removed by improperly configured systemd tmpfiles automation.
  2. then, use the thorough CryoSPARC shutdown procedure to ensure that all CryoSPARC have exited.
  3. then, try again to start CryoSPARC.
7

@wtempel when trying to ensure that the /tmp/cryosparc-supervisor-*.sock file is not subject to “improper” deletion or interference, I received a permission denied error. Please see attached. Any thoughts or suggestions to get around this is greatly appreciated.

Screenshot_20251215_135058-1
Screenshot_20251215_135058-11467×151 77.2 KB

8

Permission denied is expected when one tries to “run” (execute) the cryosparc-supervisor-*.sock file. One should not try to run/execute the cryosparc-supervisor-*.sock file.

For this purpose, the sys admin should confirm that any automated management of temporary files that may occur on the computer, such as, but not limited to

  • cron jobs
  • systemd-tmpfiles

are not configured in a way that they would automatically delete the /tmp/cryosparc-supervisor-*.sock file due to perceived “inactivity”.

9

Thank you for your prompt responses and all of your help. At this stage I am the sys admin, due to a transition. This is my first exposure to this role and I have greatly appreciated your help.

I currently performed systemd-tmpfiles --clean --dry-run to see files would be deleted and did not see the /tmp/cryosparc-supervisor-*.sock file. Following this, I did cat /usr/lib/tmpfiles.d/tmp.conf /etc/tmpfiles.d/*.conf to inspect the configuration files and found taht the entry cleans up files in /tmp older than 10 days. Could this be my issue?

dryrun
dryrun1237×667 231 KB

10

@wtempel

inspectionofconfig
inspectionofconfig1224×519 121 KB

11

Possibly. In CryoSPARC v4.7.1, one may specify alternative local paths for the CryoSPARC-related cryosparc-supervisord-*.sock and mongod-*.sock files.

  1. Identify a directory that is local (not on an nfs share) and not subject to automated clean-up. In my case, the directory /var/local fulfilled these requirements.
  2. In the directory identified in the previous step, create a directory cryosparc/. Based on the example above:
    cd /var/local/
# use sudo if needed
sudo mkdir cryosparc_sock
sudo chmod 1777 cryosparc_sock
  3. Stop CryoSPARC
  4. Append to cryosparc_master/config.sh (based on example path above):
    export CRYOSPARC_SUPERVISOR_SOCK_FILE="/var/local/cryosparc_sock/cryosparc-supervisord-$(dirname $(realpath $_) | md5sum | awk '{print $1}').sock"
export CRYOSPARC_MONGO_EXTRA_FLAGS="--unixSocketPrefix /var/local/cryosparc_sock"
  5. Start CryoSPARC and see whether cryosparcw connect [..] now works.
  6. In case you need help with a CryoSPARC configuration problem in the future, please mention that the sock files have been relocated to a custom location and therefore are not present at the presumed /tmp/ paths.