| === himcesjf_ is now known as him-cesjf | ||
| maswan | hey, we think our lineageos-builders are possibly/likely affected by https://bugzilla.kernel.org/show_bug.cgi?id=199435, the patch in comment 34 seems to have just made it into upstreams now, is it on a roadmap for a bionic update or bionic-installable hwe kernel yet? | 14:38 |
|---|---|---|
| ubot5` | bugzilla.kernel.org bug 199435 in SCSI "HPSA + P420i resetting logical Direct-Access never complete" [Normal,New] | 14:38 |
| tomreyn | maswan: this doesn't answer your question, but just in case i'll point it out anyways. i've been working with different hpe server models + generations, many of which had these cards, and while there were never any alarms, however those cards kept dying earlier than they should have. inspecting the chassis temperatures via iLOM (they have those fancy graphics indicating heat distribution across the chassis / heat sensors) showed that on | 16:48 |
| tomreyn | many (not all) models, there were hot spots (lots of red) around these very cards. and examination of bios settings showed that fans were set to some balanced mode by default. setting those to maximum cooling fixed it, the all-red spot on the sensors image went away, controllers would live longer and become more reliable, disks would die less / later, | 16:48 |
| tomreyn | unfortunately hpe never confirmed there's an issue with the default configuration there. | 16:49 |
| tomreyn | (so that's somethign you might want to check) | 16:50 |
| tomreyn | oops i mean HPE iLO, not iLOM ;-) | 16:52 |
| tomreyn | those images: https://hpeb.i.lithium.com/t5/image/serverpage/image-id/19981i3057EE35559E91A3/image-size/original | 16:55 |
| maswan | tomreyn: right, they ran rock solid as builders for a few years though, before starting to show weird-ass similar symptoms after a trusty->bionic reinstall. but thanks, will take a look at this for various hp[e] servers. | 18:14 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!