Posts
179
Following
26
Followers
383
My new server just came in today. I need to transfer the data from the old server to the new. This is going to be fun. I need to remove one of my old RAID HDs install it in the new server, get it working. Shutdown my VMs, transfer them to the new machine, and try to get them working there.

And I really have no idea what I'm doing.
1
0
2
Fun:
# echo 1 > /sys/kernel/tracing/options/func_stack_trace
# echo function > /sys/kernel/tracing/current_tracer
# time ./hackbench 5
(still waiting)
0
0
0
Is it just me, or is it harder to find motherboards, CPUs, and memory that are all compatible. I started looking for parts to build a new server and ended up just buying a refurbished Dell that had basically everything I was looking for and $500 cheaper than if I built it from parts.
0
0
2
Mastodon has made the NY Times.
https://nyti.ms/3hQbebV
0
1
1
My electric supplier notified me that my home usage is 200kwh over the average house of its size. This probably has to do with my ChromeOS build server that idles at 150w!
0
0
1
@wentasah someone suggested Husky, which works.
0
0
0
I'm trying to connect my mastodon android app to this account, but it's giving me an "Expired token" error. Anyone know how to fix it?
0
0
0

Created a new tool (WIP) That gives the output of something like this:

Total: 0.410332
  Max: 0.117048
  Min: 0.015632
  Avg: 0.041033
Count: 10
State: D
==> trace_event_raw_event_sched_switch+0xec
==> __traceiter_sched_switch+0x41
==> __schedule+0x4d3
==> schedule+0x5d
==> io_schedule+0x42
==> bit_wait_io+0xd
==> __wait_on_bit+0x4b
==> out_of_line_wait_on_bit+0x81
==> jbd2_journal_commit_transaction+0x15f1
==> kjournald2+0xa3
==> kthread+0xe9
==> ret_from_fork+0x22
Wakeup:
        ==> trace_event_raw_event_sched_wakeup_template+0x5f
        ==> try_to_wake_up+0x2e8
        ==> wake_bit_function+0x39
        ==> __wake_up_common+0x76
        ==> __wake_up_common_lock+0x6a
        ==> wake_up_bit+0x71
        ==> end_bio_bh_io_sync+0x26
        ==> clone_endio+0x103
        ==> blk_update_request+0x110
        ==> blk_mq_end_request+0x1a
        ==> blk_flush_complete_seq+0x15a
        ==> flush_end_io+0x1cf
        ==> scsi_end_request+0xdd
        ==> scsi_io_completion+0x46
        ==> blk_complete_reqs+0x40
        ==> __do_softirq+0x135

That shows the stack traces of all the places that were scheduled out in a S or D state and where it was woken up. Giving the total time of all instances of this location, as well as max, min, and average. And the number of times it was hit. There’s a surprising little number of these stacks.

Here’s the code. It depends on libtracefs 1.6.1.

0
0
4
re: [ANNOUNCE] libtracefs-1.6.0
Show content
And now 1.6.1 has been released.
0
0
1
re: [ANNOUNCE] libtracefs-1.6.0
Show content
And keeping with tradition, I found a bug as soon as I released it! 😛
1
0
0
[ANNOUNCE] libtracefs-1.6.0
Show content
libtracefs-1.6.0 has been released, with lots of cool APIs that lets you record from the raw trace files and even read it too! With examples, in the man pages.

https://git.kernel.org/pub/scm/libs/libtrace/libtracefs.git/tag/?h=libtracefs-1.6.0
1
0
1
@monsieuricon I wasn't about to say it.
0
0
0
@monsieuricon I'm guessing that it's OK to post this account to the public (like to Twitter)?
1
0
0
@monsieuricon Well, I wasn't about to tweet!
1
0
0
I guess my posts are going out. Currently working on new API for libtracefs. Hopefully it will be released by the end of the week.
0
0
0
Chirp chirp!
0
0
0
Chirp!
1
0
2
Show older