Arch Linux

SRG

Member

Registered: 2012-12-09

Posts: 54

ARCH Server "sluggish" since 2 days (through SSH)

This is driving me mad since 24h : i have a ARCH server, acting as a NAS, without a lot of things inside it.
Until now, it was very fast in all situations
.
Since one-two days, when i connect inside it through SSH, everything is weirdly sluggish : for example, if i type "ls <ENTER>", the characters will take 2-3 seconds to be printed on the screen, then i'll have the result. If it type it again, everything will be immediate. Later, another inputs will take 2-3 seconds again before being taken in account, etc.

Hence a system that seem very "sluggish". But not doing anything : no IO, no CPU being used, a lot of memory being available

I have several servers : all the other ones (sitting next to that NAS) don't have that problem (and most are also running ARCH), so i would not think about a network issue.

Any ideas ?

- I stopped some processes : same (mainly : it's only some docker containers)
- nothing special in dmesg -T, nothing in journalctl -xe
- I rebooted : same
- I downgraded the system at the date of "3 days ago" (some from kernel 6.18 to kernel 6.17 : same)
- I tried SSH-ing inside that server from another source machine : same behavior
- htop / top / ... are not showing anythings (CPU barely being issued, a lot of free memory (server has 32 GB), ...)
- iostat 2 : nothing happening (and system is on a fast NVME disk)

_lex_1234

Member

Registered: 2019-09-08

Posts: 47

Re: ARCH Server "sluggish" since 2 days (through SSH)

Did you verify the network connection? For example by pinging it, and/or ssh into the machine and ping outward from there? (and maybe speedtest / iperf from within the machine).
Also, since it acts as a NAS, is only SSH slow, but files are served at high speed? (thinking a loose ethernet cable or something like that could impact the network, sometimes without actually disconnecting).

Alex.

Last edited by _lex_1234 (2025-09-23 18:11:32)

SRG

Member

Registered: 2012-12-09

Posts: 54

Re: ARCH Server "sluggish" since 2 days (through SSH)

Mhh, maybe, indeed :
- i had checked the cable : seem ok
- i restarted the two switches around it (one being 2.5GBs) : no changes

But after your comment i just tried accessing it through Samba (from another machine) : here it's definitely also a bit sluggish (when navigating inside the folders, ...)
And when trying to download one file, it's definitely slower than usual (usually i would have been at ~130M / sec, now it's more ~50M)

No errors at first glance at network level :

enp2s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.8.139  netmask 255.255.255.0  broadcast 192.168.8.255
        inet6 fe80::ab6e:8189:adf3:90eb  prefixlen 64  scopeid 0x20<link>
        ether d8:bb:c1:16:91:5e  txqueuelen 1000  (Ethernet)
        RX packets 2606519  bytes 2879613986 (2.6 GiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 1362328  bytes 7406393095 (6.8 GiB)
        TX errors 0  dropped 3 overruns 0  carrier 0  collisions 0

_lex_1234

Member

Registered: 2019-09-08

Posts: 47

Re: ARCH Server "sluggish" since 2 days (through SSH)

Right, can you try ping from the machine (i.e. ssh into it, and ping from their either to your laptop or some other machine or the router and see if the latencies are ok, and no dropped packages?
And maybe Iperf for the networkspeed?

Last edited by _lex_1234 (2025-09-23 19:15:18)

SRG

Member

Registered: 2012-12-09

Posts: 54

Re: ARCH Server "sluggish" since 2 days (through SSH)

I don't think that server is dropping packets
Also ping seem quite ok

PING 192.168.8.5 (192.168.8.5) 56(84) bytes of data.
64 bytes from 192.168.8.5: icmp_seq=1 ttl=64 time=0.220 ms
64 bytes from 192.168.8.5: icmp_seq=3 ttl=64 time=0.224 ms
64 bytes from 192.168.8.5: icmp_seq=5 ttl=64 time=0.222 ms
64 bytes from 192.168.8.5: icmp_seq=7 ttl=64 time=0.258 ms
64 bytes from 192.168.8.5: icmp_seq=8 ttl=64 time=0.254 ms
64 bytes from 192.168.8.5: icmp_seq=10 ttl=64 time=0.222 ms
64 bytes from 192.168.8.5: icmp_seq=12 ttl=64 time=0.226 ms
64 bytes from 192.168.8.5: icmp_seq=14 ttl=64 time=0.199 ms
64 bytes from 192.168.8.5: icmp_seq=16 ttl=64 time=0.236 ms
64 bytes from 192.168.8.5: icmp_seq=18 ttl=64 time=0.251 ms
64 bytes from 192.168.8.5: icmp_seq=20 ttl=64 time=0.219 ms
64 bytes from 192.168.8.5: icmp_seq=22 ttl=64 time=0.251 ms
64 bytes from 192.168.8.5: icmp_seq=23 ttl=64 time=0.180 ms
64 bytes from 192.168.8.5: icmp_seq=25 ttl=64 time=0.217 ms
64 bytes from 192.168.8.5: icmp_seq=27 ttl=64 time=0.277 ms
64 bytes from 192.168.8.5: icmp_seq=29 ttl=64 time=0.246 ms
64 bytes from 192.168.8.5: icmp_seq=30 ttl=64 time=0.241 ms
64 bytes from 192.168.8.5: icmp_seq=32 ttl=64 time=0.210 ms
64 bytes from 192.168.8.5: icmp_seq=34 ttl=64 time=0.228 ms
64 bytes from 192.168.8.5: icmp_seq=37 ttl=64 time=0.248 ms
64 bytes from 192.168.8.5: icmp_seq=39 ttl=64 time=0.248 ms
64 bytes from 192.168.8.5: icmp_seq=41 ttl=64 time=0.228 ms
64 bytes from 192.168.8.5: icmp_seq=42 ttl=64 time=0.255 ms
64 bytes from 192.168.8.5: icmp_seq=44 ttl=64 time=0.230 ms
64 bytes from 192.168.8.5: icmp_seq=46 ttl=64 time=0.251 ms
64 bytes from 192.168.8.5: icmp_seq=47 ttl=64 time=0.256 ms
64 bytes from 192.168.8.5: icmp_seq=49 ttl=64 time=0.268 ms
(...)

Never used iperf ...

From that server (as client) against another machine (as iperf server) (the two being with 2.5GB)

21:48 root@saturne ~# iperf -c  192.168.8.5
------------------------------------------------------------
Client connecting to 192.168.8.5, TCP port 5001
TCP window size: 16.0 KByte (default)
------------------------------------------------------------
[  1] local 192.168.8.139 port 40220 connected with 192.168.8.5 port 5001
[ ID] Interval       Transfer     Bandwidth
[  1] 0.0000-10.9431 sec   877 MBytes   672 Mbits/sec

Between two not related machines (= that "sluggish" server is not involved) (but here one is 2.5GB, the other one is 1GB) :

21:51 root@jupiter ~# iperf -c 192.168.8.5
------------------------------------------------------------
Client connecting to 192.168.8.5, TCP port 5001
TCP window size: 85.0 KByte (default)
------------------------------------------------------------
[  1] local 192.168.8.4 port 41598 connected with 192.168.8.5 port 5001
[ ID] Interval       Transfer     Bandwidth
[  1] 0.0000-10.0251 sec   995 MBytes   833 Mbits/sec

(so even a bit better whereas here i'm capped at 1Gbps at network level on one (old) machine)

When unplugging/replugging the network cable (on the problematic server), i correctly see the proper speed :

[Tue Sep 23 19:01:58 2025] r8169 0000:02:00.0 enp2s0: Link is Up - 2.5Gbps/Full - flow control rx/tx

Last edited by SRG (2025-09-23 19:58:43)

SRG

Member

Registered: 2012-12-09

Posts: 54

Re: ARCH Server "sluggish" since 2 days (through SSH)

Ok, so i changed the RJ45 cable (that was "untouched" since a long time ... so no reasons to suddenly not work anymore !) and with a new cable, it's working fine ... weird that it was "sluggish" without real disconnects ...

_lex_1234

Member

Registered: 2019-09-08

Posts: 47

Re: ARCH Server "sluggish" since 2 days (through SSH)

Ok, so i changed the RJ45 cable (that was "untouched" since a long time ... so no reasons to suddenly not work anymore !) and with a new cable, it's working fine ... weird that it was "sluggish" without real disconnects ...

Right, yes, it is a little unsatisfatory that you still don't know for sure, but good that it works now!
I had similar unclear network problems once, related to flacky ethernet cables/connections, but I could at least also see the problems in bandwidth measurements :-)