Taking long time to launch a container

LXD
1

Hello,

I am running lxd 5.9 on debian 11.0 host machine. While creating a container, it is taking long time to create (seems like stuck).

root@monitor:~# lxc launch images:debian/bullseye test
Creating test

can you help me on this ?

Thanks

2

Could be a network issue if it’s the first time a particular image is used.

You can check what’s going on with lxc operation list but there’s a good chance you’re stuck in image download.

3

@stgraber thanks for the fast response

the image is already downloaded on the host

root@monitor:~# lxc image list
+-------+--------------+--------+----------------------------------------+--------------+-----------+---------+-----------------------------+
| ALIAS | FINGERPRINT  | PUBLIC |              DESCRIPTION               | ARCHITECTURE |   TYPE    |  SIZE   |         UPLOAD DATE         |
+-------+--------------+--------+----------------------------------------+--------------+-----------+---------+-----------------------------+
|       | 93d985be253b | no     | Debian bullseye amd64 (20230104_05:25) | x86_64       | CONTAINER | 83.54MB | Jan 5, 2023 at 3:29am (UTC) |
+-------+--------------+--------+----------------------------------------+--------------+-----------+---------+-----------------------------+

it taking about 3-4mins to create and delete a container.
also checking lxc operation list, seems it shows normal

root@monitor:~# lxc operation list
+--------------------------------------+------+-------------------+---------+------------+----------------------+
|                  ID                  | TYPE |    DESCRIPTION    | STATUS  | CANCELABLE |       CREATED        |
+--------------------------------------+------+-------------------+---------+------------+----------------------+
| 2e3e71df-2b7c-4d40-a7c7-e3714b7009bc | TASK | Creating instance | RUNNING | NO         | 2023/01/05 14:07 UTC |
+--------------------------------------+------+-------------------+---------+------------+----------------------+

root@monitor:~# lxc operation list
+--------------------------------------+------+-------------------+---------+------------+----------------------+
|                  ID                  | TYPE |    DESCRIPTION    | STATUS  | CANCELABLE |       CREATED        |
+--------------------------------------+------+-------------------+---------+------------+----------------------+
| 10625d7a-5475-416c-8ba5-6565790d4233 | TASK | Stopping instance | RUNNING | NO         | 2023/01/05 14:12 UTC |
+--------------------------------------+------+-------------------+---------+------------+----------------------+

Is there any method to find and troubleshoot the issue?

Thanks

4

Hmm, I’d probably look into the usual suspects:

  • dmesg for anything kernel related
  • ps fauxww and top to see what’s using CPU/memory when things are slow

You can also run lxc monitor --pretty in another terminal which will get you LXD debugging logs.
This may give more details as far as what’s taking so long.

5

Benchmark your system first to see if everything checks out (what filesystem are you using?)

#Sequential write test
sync;fio --randrepeat=1 --ioengine=libaio --direct=1 --name=test --filename=test --bs=4M --size=4G --readwrite=write --ramp_time=4 numjobs=$(nproc)

#Sequential Read test
sync;fio --randrepeat=1 --ioengine=libaio --direct=1 --name=test --filename=test --bs=4M --size=4G --readwrite=read --ramp_time=4 numjobs=$(nproc)

Run an iperf test as well while you’re at it (iperf not iperf3)

iperf -s

iperf -c localhost -P $(nproc)

then use lxd-benchmarking tool