-
hyc
btw, I had also restarted sync from scratch, using ethernet cable instead of wifi, and didn't hit this issue for at least 1M blocks
-
hyc
so there was definitely a timing element
-
moneromooo
1M sounds similar to how much I had to wait before seeing it.
-
moneromooo
Though I did the first 125k repeatedly.
-
wfaressuissia[m]
after runtime debugging vtable is correct now
-
wfaressuissia[m]
realized how to print content of boost::unordered_map , printed for all connections m_remote_address, turned out all of them ok
-
moneromooo
But terminate is still called ?
-
wfaressuissia[m]
let me share backtrace
-
wfaressuissia[m]
any service to share online vimdiff ?
-
wfaressuissia[m]
pasted to
paste.dev.centos.org/view/raw/6dd8ff2f but it triggers some internal error
-
wfaressuissia[m]
in their website
-
hyc
try paste.debian.net
-
wfaressuissia[m]
size limitted
-
hyc
huh. gcc 10.2.0 segfaults while compiing the tree
-
moneromooo
paste.ubuntu.com works for longer.
-
wfaressuissia[m]
there is raw url for anonymous, ok for you ?
-
moneromooo
I can save and edit.
-
hyc
this run is looking good, still on ethernet. I'll switch back to wifi.
-
hyc
heh. ok, had another mystery disconnect, but this time I had a kernel error msg, ethernet driver flaked.
-
hyc
had to ifconfig down/up the interface to resume
-
hyc
sometimes it's a mirale anything works at all
-
hyc
miracle
-
selsta
.merge+ 7269 7268
-
xmr-pr
Added
-
hyc
hmm. 2021-01-03 02:12:00.689 W Failed to commit a transaction to the db: File too large
-
hyc
auto resize isn't triggering soon enough?
-
hyc
2021-01-03 02:12:00.6725621319217
-
hyc
I HEIGHT 1227287, difficulty:
-
hyc
5621319217
-
hyc
and looks like a segv in logger while stopping monerod
paste.debian.net/1179498
-
hyc
yeah I'm stuck at height 1227327, auto resize not kicking in
-
selsta
even after restarting?
-
hyc
yep still no joy
-
hyc
wiping it and starting again
-
moneromooo
No logs of it ?
-
hyc
I have the bitmonero.log but there's nothing useful or unusual in it
-
hyc
no other error messages
-
selsta
.merges
-
xmr-pr
7248 7261 7262 7263 7264 7268 7269 7271 7272
-
selsta
vtnerd: please take a look at
monero-project/monero #7248 again. mooo implemented your suggestions
-
moneromooo
(some)
-
Mmmmmmmmmm
Why do the Mac and Windows CI checks fail?
-
hyc
... so, back on wifi. I guess the network speed actually does make a measurable difference. CPU use was around 100% on ethernet. back down in the 80% range on wifi.
-
hyc
getting close. 1030640, no errors yet
-
hyc
just blew past it. oh well. sync still going fine, no errors
-
M5M400
just wanted to say that v0.17.1.8 is the first version since 0.17.x that survived 3 days without segfault
-
M5M400
on node.supportxmr.com
-
hyc
it has definitely not survived without segfault for me, doing sync from scratch
-
hyc
but the recent patches seem to be behaving OK
-
M5M400
weird. running it on 6 nodes and not a single issue since release
-
M5M400
even using dns block list that is supposed to cause problems
-
hyc
well, my VPS node is also running it fine, but that was just continuing its existing blockchain DB
-
M5M400
ah, k. I didn't sync from scratch
-
hyc
I'm now working under the assumption my SSD is flaking out
-
hyc
which is a bit annoying, but the last time I used this system was like a year ago, and it had no problems back then
-
hyc
still, SSDs have finite life spans
-
sech1
SSD without a power for a year? It might loose some data
-
hyc
yeah
-
hyc
it appears to be corrupting data now
-
hyc
well, I don't even know that yet. nothing in kernel logs.
-
hyc
but I get some DB errors during sync
-
hyc
definitely seems to be bad blocks on my SSD. Now every time monerod crashes I rename data.mdb so its sectors don't get reused.
-
hyc
and the sync gets further when I restart it
-
gingeropolous
aight im syncing from scratch across 7 different shitty ssds in raid
-
gingeropolous
log 4 baby. gonna catch em all
-
gingeropolous
of course its not debug. :(
-
gingeropolous
well i'll do it again
-
sech1
Is it RAID0?
-
gingeropolous
yeah the one with no redundancy
-
gingeropolous
well it might be the new raidz, i forget now
-
ErCiccione[m]
I'm syncing a stagenet node that seems to be banning nodes. Sounds weird.
-
ErCiccione[m]
6 banned within a couple of hours
-
moneromooo
Did you keep logs ?
-
ErCiccione[m]
i did not.
-
ErCiccione[m]
I'm giving it some proper settings now. i simply run it with monerod --stagenet
-
selsta
I assume that’s the same false positives we have why syncing on main chain
-
gingeropolous
hrm, sync got killed
-
gingeropolous
guess i'll compile debug
-
selsta
gingeropolous: which version
-
gingeropolous
v0.17.1.8
-
gingeropolous
syncing of of an exclusive local node
-
selsta
gingeropolous: if you want to compile yourself, release-v0.17 + 7248 + 7262 + 7264 + 7266 + 7269 + 7272 will be v0.17.1.9 most likely
-
gingeropolous
ok, so checkout v0.17.1.8 and these PRs?
-
gingeropolous
7266 gave me a conflict
-
gingeropolous
using this approach: git pull origin pull/7266/head
-
gingeropolous
, and doing them in sequence adding each PR starting at 7248
-
selsta
release-v0.17 + 7248 + 7262 + 7264 + 7267 + 7269 + 7272 will be v0.17.1.9 most likely
-
selsta
yea had a typo
-
selsta
git checkout release-v0.17
-
selsta
git pull origin pull/7248/head
-
selsta
and the same for others
-
gingeropolous
resolve my current index my ass. ima nuke u folder
-
sech1
rm -rf monero
-
gingeropolous
thanks for your patience selsta
-
gingeropolous
hrm, i added --add-exclusive-node yet i still have 5(out)+8(in)
-
gingeropolous
i wonder if running in gdb made parsing those commands odd. though it did goto the data directory i specified
-
gingeropolous
did new peerlist stuff cause an override of exclusive node startup flags?
-
iDunk
gdb --args /path/to/monerod --add-exclusive-node a.b.c.d
-
gingeropolous
ah, i did gdb ./monerod , then ran r --(insert flags)
-
sgp_
-
mj-xmr
Just completed unity build for both core tests and unit tests @ 7217
-
mj-xmr
Good night