Nimbostratus
NimbostratusHoolio,
Thanks for your reply. I'm really just trying to better understand what is utilizing the memory on the box. I'm not looking at the graphs, I'm looking at the memory statistics in the GUI which seems to be the same information shown by running the tmsh show sys memory command.
To elaborate, I have multiple F5 pairs with similar configurations and in one of my environments the memory utilization seems a little high compared to the load. I'm interested in what each of the memory pools (as it is labeled in the GUI) / SubSystems (as it is labeled in the TMSH output) are responsible for.
For instance:
SubSystem Alloc (bytes) Max Alloc (bytes) Object Size
filter (variable) 1.8G 1.9G 1
ssl_cn 545.4M 571.0M 1064
ssl_session 214.8M 226.4M 224
tm_header_cache_entry_slab 206.4M 216.0M 384
They seem a little self-explanatory, but I don't want to make any assumptions.
John
NimbostratusI would also like to know this. I have a memory pool that is increasing in size daily and feel i could have hit a memory leak issue. I would like to know what the memory pool named 'persist' is, there is a separate memory pool called persistence which is not expanding in size. It would be handy to know what all memory pools were to help identify causes of leaks/high usage
I am running 11.5.4 Hot Fix 2 and there are no listed bugs i can find for this
Memory Pool Name Allocated Max Allocated Object Size
persist 5.9G 5.9G 160
persistence 256.0M 256.0M 1
The persist memory pool is steadily growing about 100MB per day whereas persistence is stable
Edit: My issue was a bug 562566-3 which causes the timeout of persistence entries to be reset to 0, so it holds onto all persistence entries forever. We fixed it by first rebooting, then later upgrading to v12. The cause is supposedly an HA flap however we didn't have any flaps that i know of, our devices just started holding onto persistence entries until we rebooted. It didn't happen on any of my other v11 pairs so something did kick it off on this pair. It would still be nice to get a page that lists all the memory pools and what creates/uses/clears them so we can identify issues more quickly