Chapter 3. Troubleshooting networking issues

Procedure

1. Installing the net-tools and telnet packages can help when troubleshooting network issues that can occur in a Ceph storage cluster:

   Example

   [root@host01 ~]# dnf install net-tools
   [root@host01 ~]# dnf install telnet

2. Log into the cephadm shell and verify that the public_network parameters in the Ceph configuration file include the correct values:

   Example

   [ceph: root@host01 /]# cat /etc/ceph/ceph.conf
   # minimal ceph.conf for 57bddb48-ee04-11eb-9962-001a4a000672
   [global]
   	fsid = 57bddb48-ee04-11eb-9962-001a4a000672
   	mon_host = [v2:10.74.249.26:3300/0,v1:10.74.249.26:6789/0] [v2:10.74.249.163:3300/0,v1:10.74.249.163:6789/0] [v2:10.74.254.129:3300/0,v1:10.74.254.129:6789/0]
   [mon.host01]
   public network = 10.74.248.0/21

3. Exit the shell and verify that the network interfaces are up:

   Example

   [root@host01 ~]# ip link list
   1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
       link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
   2: ens3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
       link/ether 00:1a:4a:00:06:72 brd ff:ff:ff:ff:ff:ff

4. Verify that the Ceph nodes are able to reach each other using their short host names. Verify this on each node in the storage cluster:

   Syntax

   ping SHORT_HOST_NAME

   Example

   [root@host01 ~]# ping host02

5. If you use a firewall, ensure that Ceph nodes are able to reach each other on their appropriate ports. The firewall-cmd and telnet tools can validate the port status, and if the port is open respectively:

   Syntax

   firewall-cmd --info-zone=ZONE
   telnet IP_ADDRESS PORT

   Example

   [root@host01 ~]# firewall-cmd --info-zone=public
   public (active)
     target: default
     icmp-block-inversion: no
     interfaces: ens3
     sources:
     services: ceph ceph-mon cockpit dhcpv6-client ssh
     ports: 9283/tcp 8443/tcp 9093/tcp 9094/tcp 3000/tcp 9100/tcp 9095/tcp
     protocols:
     masquerade: no
     forward-ports:
     source-ports:
     icmp-blocks:
     rich rules:
   
   [root@host01 ~]# telnet 192.168.0.22 9100

6. Verify that there are no errors on the interface counters. Verify that the network connectivity between nodes has expected latency, and that there is no packet loss.

   1. Using the ethtool command:

      Syntax

      ethtool -S INTERFACE

      Example

      [root@host01 ~]# ethtool -S ens3 | grep errors
      NIC statistics:
           rx_fcs_errors: 0
           rx_align_errors: 0
           rx_frame_too_long_errors: 0
           rx_in_length_errors: 0
           rx_out_length_errors: 0
           tx_mac_errors: 0
           tx_carrier_sense_errors: 0
           tx_errors: 0
           rx_errors: 0

   2. Using the ifconfig command:

      Example

      [root@host01 ~]# ifconfig
      ens3: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
              inet 10.74.249.26  netmask 255.255.248.0  broadcast 10.74.255.255
              inet6 fe80::21a:4aff:fe00:672  prefixlen 64  scopeid 0x20<link>
              inet6 2620:52:0:4af8:21a:4aff:fe00:672  prefixlen 64  scopeid 0x0<global>
              ether 00:1a:4a:00:06:72  txqueuelen 1000  (Ethernet)
              RX packets 150549316  bytes 56759897541 (52.8 GiB)
              RX errors 0  dropped 176924  overruns 0  frame 0
              TX packets 55584046  bytes 62111365424 (57.8 GiB)
              TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
      
      lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536
              inet 127.0.0.1  netmask 255.0.0.0
              inet6 ::1  prefixlen 128  scopeid 0x10<host>
              loop  txqueuelen 1000  (Local Loopback)
              RX packets 9373290  bytes 16044697815 (14.9 GiB)
              RX errors 0  dropped 0  overruns 0  frame 0
              TX packets 9373290  bytes 16044697815 (14.9 GiB)
              TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

   3. Using the netstat command:

      Example

      [root@host01 ~]# netstat -ai
      Kernel Interface table
      Iface             MTU    RX-OK RX-ERR RX-DRP RX-OVR    TX-OK TX-ERR TX-DRP TX-OVR Flg
      ens3             1500 311847720      0 364903 0      114341918      0      0      0 BMRU
      lo              65536 19577001      0      0 0      19577001      0      0      0 LRU

7. For performance issues, in addition to the latency checks and to verify the network bandwidth between all nodes of the storage cluster, use the iperf3 tool. The iperf3 tool does a simple point-to-point network bandwidth test between a server and a client.

   1. Install the iperf3 package on the Red Hat Ceph Storage nodes you want to check the bandwidth:

      Example

      [root@host01 ~]# dnf install iperf3

   2. On a Red Hat Ceph Storage node, start the iperf3 server:

      Example

      [root@host01 ~]# iperf3 -s
      -----------------------------------------------------------
      Server listening on 5201
      -----------------------------------------------------------

      Note

      The default port is 5201, but can be set using the -P command argument.

   3. On a different Red Hat Ceph Storage node, start the iperf3 client:

      Example

      [root@host02 ~]# iperf3 -c mon
      Connecting to host mon, port 5201
      [  4] local xx.x.xxx.xx port 52270 connected to xx.x.xxx.xx port 5201
      [ ID] Interval           Transfer     Bandwidth       Retr  Cwnd
      [  4]   0.00-1.00   sec   114 MBytes   954 Mbits/sec    0    409 KBytes
      [  4]   1.00-2.00   sec   113 MBytes   945 Mbits/sec    0    409 KBytes
      [  4]   2.00-3.00   sec   112 MBytes   943 Mbits/sec    0    454 KBytes
      [  4]   3.00-4.00   sec   112 MBytes   941 Mbits/sec    0    471 KBytes
      [  4]   4.00-5.00   sec   112 MBytes   940 Mbits/sec    0    471 KBytes
      [  4]   5.00-6.00   sec   113 MBytes   945 Mbits/sec    0    471 KBytes
      [  4]   6.00-7.00   sec   112 MBytes   937 Mbits/sec    0    488 KBytes
      [  4]   7.00-8.00   sec   113 MBytes   947 Mbits/sec    0    520 KBytes
      [  4]   8.00-9.00   sec   112 MBytes   939 Mbits/sec    0    520 KBytes
      [  4]   9.00-10.00  sec   112 MBytes   939 Mbits/sec    0    520 KBytes
      - - - - - - - - - - - - - - - - - - - - - - - - -
      [ ID] Interval           Transfer     Bandwidth       Retr
      [  4]   0.00-10.00  sec  1.10 GBytes   943 Mbits/sec    0             sender
      [  4]   0.00-10.00  sec  1.10 GBytes   941 Mbits/sec                  receiver
      
      iperf Done.

      This output shows a network bandwidth of 1.1 Gbits/second between the Red Hat Ceph Storage nodes, along with no retransmissions (Retr) during the test.

      Red Hat recommends you validate the network bandwidth between all the nodes in the storage cluster.

8. Ensure that all nodes have the same network interconnect speed. Slower attached nodes might slow down the faster connected ones. Also, ensure that the inter switch links can handle the aggregated bandwidth of the attached nodes:

   Syntax

   ethtool INTERFACE

   Example

   [root@host01 ~]# ethtool ens3
   Settings for ens3:
   Supported ports: [ TP ]
   Supported link modes:   10baseT/Half 10baseT/Full
                           100baseT/Half 100baseT/Full
                           1000baseT/Half 1000baseT/Full
   Supported pause frame use: No
   Supports auto-negotiation: Yes
   Supported FEC modes: Not reported
   Advertised link modes:  10baseT/Half 10baseT/Full
                           100baseT/Half 100baseT/Full
                           1000baseT/Half 1000baseT/Full
   Advertised pause frame use: Symmetric
   Advertised auto-negotiation: Yes
   Advertised FEC modes: Not reported
   Link partner advertised link modes:  10baseT/Half 10baseT/Full
                                        100baseT/Half 100baseT/Full
                                        1000baseT/Full
   Link partner advertised pause frame use: Symmetric
   Link partner advertised auto-negotiation: Yes
   Link partner advertised FEC modes: Not reported
   Speed: 1000Mb/s 1
   Duplex: Full 2
   Port: Twisted Pair
   PHYAD: 1
   Transceiver: internal
   Auto-negotiation: on
   MDI-X: off
   Supports Wake-on: g
   Wake-on: d
   Current message level: 0x000000ff (255)
          drv probe link timer ifdown ifup rx_err tx_err
   Link detected: yes 3

Procedure

1. Verify that the chronyd daemon is running on the Ceph Monitor hosts:

   Example

   [root@mon ~]# systemctl status chronyd

2. If chronyd is not running, enable and start it:

   Example

   [root@mon ~]# systemctl enable chronyd
   [root@mon ~]# systemctl start chronyd

3. Ensure that chronyd is synchronizing the clocks correctly:

   Example

   [root@mon ~]# chronyc sources
   [root@mon ~]# chronyc sourcestats
   [root@mon ~]# chronyc tracking