a year ago
I recently had to migrate a project to digital ocean due to slow performance.
My question is not really project specific. I have put a few "trial projects" on railway and noticed they are not very performant vs when i deploy the identical dockerimage to minimum spec digital ocean for example.
Is there some way I can pay for a better or dedicated host?
The plan says "32 GB RAM / 32 vCPU per service".
I'm not taking a dig at railway here, i genuinly want a solution so i can start use the platform more.
0 Replies
Hey, I noticed this as well today. Just wanted to say Railway is not normally like this. I have had extremely bad performance with it the past several hours, so I think you just hit some unlucky timing. They might be getting DDoSed (happened in the past), overloaded, or similar
but i dont even understand how that happens, cpu shouldn't be oversubscribed. and if someone gets ddos they can use up their CPU shares. and as long as there is enough on the host left for me then im good.
and the way it reads. per pro account gets allocated.
32 GB RAM / 32 vCPU per service
and my thing that was running slow was just a simple graphql api with a database.
a year ago
what cpu do you get on your do droplet?
a year ago
im talking about the actual cpu, not vcpu
/myrounds # cat /proc/cpuinfo
processor : 0
vendorid : GenuineIntel cpu family : 6 model : 106 model name : unknown stepping : unknown cpu MHz : 2593.904 fpu : yes fpuexception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx pdpe1gb rdtscp lm pni pclmulqdq vmx ssse3 fma cx16 pcid sse41 sse42 x2apic movbe popcnt tscdeadlinetimer aes xsave avx f16c rdrand hypervisor lahflm abm 3dnowprefetch fsgsbase bmi1 avx2 smep bmi2 erms invpcid avx512f avx512dq rdseed adx smap clwb avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 avx512vbmi umip pku ospke avx512vbmi2 gfni vaes vpclmulqdq avx512vnni avx512bitalg avx512vpopcntdq bogomips : 2593.90 clflush size : 64 cachealignment : 64
address sizes : 4
a year ago
root@railway:/# cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 85
model name : Intel(R) Xeon(R) CPU @ 2.80GHz
stepping : 7
microcode : 0xffffffff
cpu MHz : 2799.998
cache size : 33792 KB
physical id : 0
siblings : 16
core id : 0
cpu cores : 8
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq vmx ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap clflushopt clwb avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves arat avx512_vnni md_clear arch_capabilities
vmx flags : vnmi preemption_timer invvpid ept_x_only ept_ad flexpriority tsc_offset vtpr mtf vapic ept vpid unrestricted_guest vapic_reg shadow_vmcs
bugs : spectre_v1 spectre_v2 spec_store_bypass mds swapgs taa mmio_stale_data retbleed eibrs_pbrsb gds
bogomips : 5599.99
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtuala year ago
technically close wise, you already do
a year ago
but no, until they move to bare metal they dont have different hardware to run your service on
a year ago
but i dont think this is a hardware bottleneck tbh