-
Notifications
You must be signed in to change notification settings - Fork 2k
Out of memory error #440
Comments
What kind of machine are you running this on? Are there any memory limits set on your docker? |
8 CPU thread (4 cores + HT) and 24GB RAM... would expect it to be enough. However, I do not see RAM usage increase all that much, maybe 1GB or so. I have not manually set any limits on docker and I generally have a lot more and larger containers running. Unless the NodeJS container is 32bit and can only support up to 3.5GB of RAM, I doubt the problem is in my machine. Any place else where I should be looking? |
I got this too a few days back |
me too, and I don't use docker
|
same with i5 and 16 GB <--- Last few GCs ---> [23401:0x2fdfb00] 420607 ms: Mark-sweep 1412.0 (1468.5) -> 1411.8 (1455.5) MB, 1723.5 / 1.5 ms (+ 0.0 ms in 0 steps since start of marking, biggest step 0.0 ms, walltime since start of marking 1724 ms) last resort GC in old space requested <--- JS stacktrace ---> ==== JS stack trace ========================================= Security context: 0x117c43a25ee1 FATAL ERROR: CALL_AND_RETRY_LAST Allocation failed - JavaScript heap out of memory |
I also have an I5 and 16 gb ram and i get this. Crashes my computer. |
Just got this today:
|
Same error here. First darwin wouldn't give any good results, showed currency_capital error so I tried running the sim myself. It became very slow minutes before the error was shown, not usual speedy sim whizzing by. This one ground to a halt. 2 things changed from previous generations.
Resulted from the commandzenbot.bat sim gdax.LTC-USD --period_length=241m --min_periods=40 --markdown_buy_pct=-0.9817790501161023 --markup_sell_pct=0.33335068998603434 --order_type=maker --sell_stop_pct=30 --buy_stop_pct=8 --profit_stop_enable_pct=9 --profit_stop_pct=9 --trend_ema=13 --oversold_rsi_periods=19 --oversold_rsi=20 --backtester_generation=1 --strategy=trend_ema --maxCores=2 --start=201712111340 --end=201801160100 --asset_capital=4 --currency_capital=1000 --filename=none Error message` [1756:000002080C5C0170] 191100 ms: Mark-sweep 1405.1 (1433.3) -> 1405.1 (1433.3) MB, 1174.2 / 0.0 ms last resort GC in old space requested <--- JS stacktrace ---> ==== JS stack trace ========================================= Security context: 0000019DD39A5EE1 FATAL ERROR: CALL_AND_RETRY_LAST Allocation failed - JavaScript heap out of memory |
Also have this error ... The command runs with normal behaviour for a few data-days (this ingestion take about 10 minutes) and then is slows to a crawl before crashing back to the shell. Resulted from command:
Starts out as:
until it starts slowing down to a crawl
and it crashes Error message
|
Memory workarounds:
|
@glennfu 👍 that works like a charm :-) ... |
So i ran in to this also on a raspberry pi 4. Thinking it was a pice of crap i dusted of a old ibm server with loads of memory. Happily installing Ubuntu thinking to beat that error! Aand i get the same one even if it was running on less than 10% of total memory... Can be fixed easily enough by running zenbot with this command: Node <8.0.0 it should be: That will let it use more than the default set limit of memory. Zenbot seems to be quite happy running on 1.8Gb max memory in general, but you can set it to what you want. #increase to 2gb #increase to 3gb #increase to 4gb #increase to 5gb #increase to 6gb #increase to 7gb #increase to 8gb For global Node memory increase just type:
and so on. Maybe someone will find this useful :) |
@DeviaVir Think you can close this bug :) |
Might still be useful to document this https://github.com/DeviaVir/zenbot/tree/unstable/docs |
* Set stdin_open to true in docker-compose.yml - allows interactive trading commands when using docker-compose * Add docu for "JavaScript heap out of memory" Discussed in #440
So I got the following stack trace after I had been trading for a little while:
At this point, it seems zenbot no longer wants to trade this pair, as it keeps stuck on:
I'm running the code in Docker containers and another trading bot is running just fine (in another container). Any hints on what's going on here or what I should be doing to clean this up? I'm using the trust_distrust strategy for this bot.
Any help is appreciated!
The text was updated successfully, but these errors were encountered: