Fatal Config Errror - memory limit not set

1

Hi,

I would like to report an error during install using frappe_setup.sh.
The install is clean and succesfully, the 3 passwords also displayed. ( not using option --production in the command)

Install in Centos 6.5 32 bit machine.

When i start the bench ( # bench start), i got the following error message:

13:15:04 system | workerbeat.1 started (pid=11372)
13:15:04 system | worker.1 started (pid=11374)
13:15:04 system | web.1 started (pid=11375)
13:15:04 system | async_worker.1 started (pid=11377)
13:15:04 system | socketio.1 started (pid=11379)
13:15:04 system | redis_async_broker.1 started (pid=11381)
13:15:04 system | redis_cache.1 started (pid=11384)
13:15:04 system | longjob_worker.1 started (pid=11386)
13:15:04 redis_cache.1 |
13:15:04 redis_cache.1 | *** FATAL CONFIG FILE ERROR ***
13:15:04 redis_cache.1 | Reading the configuration file, at line 6
13:15:04 redis_cache.1 | >>> ‘save “”’
13:15:04 redis_cache.1 | Bad directive or wrong number of arguments
13:15:04 system | redis_cache.1 stopped (rc=1)
13:15:04 system | sending SIGTERM to workerbeat.1 (pid 11372)
13:15:04 system | sending SIGTERM to worker.1 (pid 11374)
13:15:04 system | sending SIGTERM to web.1 (pid 11375)
13:15:04 system | sending SIGTERM to async_worker.1 (pid 11377)
13:15:04 system | sending SIGTERM to socketio.1 (pid 11379)
13:15:04 system | sending SIGTERM to redis_async_broker.1 (pid 11381)
13:15:04 system | sending SIGTERM to longjob_worker.1 (pid 11386)
13:15:04 system | socketio.1 stopped (rc=-15)
13:15:04 system | async_worker.1 stopped (rc=-15)
13:15:04 system | longjob_worker.1 stopped (rc=-15)
13:15:04 system | workerbeat.1 stopped (rc=-15)
13:15:04 system | web.1 stopped (rc=-15)
13:15:04 redis_async_broker.1 | [11381] 16 Feb 13:15:04 # Received SIGTERM, scheduling shutdown…
13:15:04 redis_async_broker.1 | [11381] 16 Feb 13:15:04 # Warning: 32 bit instance detected but no memory limit set. Setting 3.5 GB maxmemory limit with ‘noeviction’ policy now.
13:15:04 system | worker.1 stopped (rc=-15)
13:15:04 redis_async_broker.1 | [11381] 16 Feb 13:15:04 * Server started, Redis version 2.4.10
13:15:04 redis_async_broker.1 | [11381] 16 Feb 13:15:04 * The server is now ready to accept connections on port 12311
13:15:05 redis_async_broker.1 | [11381] 16 Feb 13:15:05 # User requested shutdown…
13:15:05 redis_async_broker.1 | [11381] 16 Feb 13:15:05 # Redis is now ready to exit, bye bye…
13:15:05 system | redis_async_broker.1 stopped (rc=0)

the first time error, give me direction to add ‘vm.overcommit_memory=1’ into /etc/sysctl.conf

i do follw by adding it in the last line of the file. Then reboot the machine. (it seem hard to get up after i add the line above, 2-3 times booting try)

when it boot succesfully and start again, the above error displayed.

Please help, where to and how to set the directive rightly or solve this error.

Thank you,
Bun Hin

2

Are you using vagrant?

3

remove the save line in config/redis_cache.conf

4

Hi Anand,
Thank you for sugested fix, i will try it soon, and will inform the outcome.
Now i am still away and can not reach the box

Greetings,
Bun Hin

5

Hi Anand,

Ok. I find the config/redis_cache. conf file
delete the line: save “”
and now can start the bench.

but while it it is running it keeps give report about the status and not back to the command prompt.
i can get back to command prompt by pressing Ctrl+C but it is stopping the server.
How to back to command prompt without stopping then?

Thank you
Bun Hin

6

That’s the development setup. Server runs till you press ctrl+c. Helps you debug and restart processes as and when required.

For production deployment, see: GitHub - frappe/bench: CLI to manage Multi-tenant deployments for Frappe apps