Ikamva le-AIOps: Ukudibanisa i-AI yoLawulo lwe-IT olusebenzayo | B4N1

Ikamva le-AIOps: Ukudibanisa i-AI yoLawulo lwe-IT olusebenzayo

9 months ago · b4n1 Read time: 3 min

Intshayelelo

Ukukhula ngokukhawuleza kwethekhnoloji kukhokelele ekukhuleni okucacileyo kububunzima beenkqubo ze-IT zanamhlanje. Ngenxa yoko, amaqela e-IT ajongene nemingeni engazange ibonwe ngaphambili ekulawuleni nasekugcineni ezi nkqubo, ezikhokelela ekunciphiseni ixesha lokunciphisa, ukunciphisa imveliso, kunye neendleko eziphezulu zokusebenza. Ukujongana nale mingeni, ingcamango ye-Artificial Intelligence ye-IT Operations (AIOps) iye yavela njengomtshintshi womdlalo. I-AIOps iphakamisa ukufundwa koomatshini kunye ne-AI ukuguqula ukuhamba komsebenzi kwe-DevOps, ivumela ulawulo lwe-IT olusebenzayo, uhlalutyo lokuxela kwangaphambili, kunye nokuzenzekelayo.

Kweli nqaku, siza kuphonononga ikamva le-AIOps, izibonelelo zayo, kunye nezicwangciso zokuphumeza. Siza kuphinda sihlolisise kwizifundo zangempela zehlabathi kwaye sinikeze ibhloko yekhowudi enzima ukubonisa ukusetyenziswa kwe-AIOps kulawulo lwe-IT.

Yintoni i-AIOps?

I-AIOps yintsimi ephumayo edibanisa amandla obuhlakani bokufakelwa (AI) kunye nokufunda komatshini (ML) kunye nokusebenza kwe-IT (ITOps) ukuphucula ukusebenza kakuhle, ukusebenza, kunye nokuthembeka kweenkonzo ze-IT. Iiplatifti ze-AIOps zisebenzisa uhlalutyo lwedatha, imodeli yokuqikelela, kunye ne-automation ukuchonga nokusombulula imiba ye-IT ngaphambi kokuba ibe yingxaki enkulu.

Ii-AIOps zinokusetyenziswa kwiindawo ezahlukeneyo zolawulo lwe-IT, kubandakanya:

  1. I-Analytics Predictive: Iiplatifomu ze-AIOps zisebenzisa i-algorithms yokufunda ngomatshini ukuhlalutya idatha yembali kunye nokuqikelela imiba enokubakho ye-IT, eyenza ukugcinwa ngokukhawuleza kunye nokunciphisa ixesha lokuphumla.
  2. **Ulawulo lwezehlo **: Ii-AIOps ziyakwazi ukubona ngokuzenzekelayo kwaye zisombulule iziganeko ze-IT, ukunciphisa ixesha eliqhelekileyo lokubona (MTTD) kunye nexesha elichanekileyo lokusombulula (MTTR).
  3. I-Automation: Ii-AIOps zinokwenza ngokuzenzekelayo imisebenzi yesiqhelo, njengolawulo loqwalaselo, ulawulo lweepatch, kunye ne-backup kunye nokubuyisela.

Izibonelelo ze-AIOps

Izibonelelo ze-AIOps zininzi kwaye zibalulekile:

  1. Ukusebenza okuPhuculweyo: I-AIOps izenzela imisebenzi yesiqhelo, ikhulula abasebenzi be-IT ukuba bagxininise kumalinge acwangcisiweyo.
  2. Ukunyaniseka okuNyenziwe: I-AIOps iqikelela kwaye inqanda imiba ye-IT, ukunciphisa ixesha lokuphumla kunye nokuphucula ukuthembeka kwenkqubo iyonke.
  3. **Ukonyuka kweMveliso **: I-AIOps yenza ukuba amaqela e-IT aphendule ngokukhawuleza kwiziganeko, ukunciphisa ixesha eliqhelekileyo lokubona (MTTD) kunye nexesha elichanekileyo lokusombulula (MTTR).
  4. **Ukwenziwa kweZigqibo eziNgcono **: I-AIOps ibonelela ngamaqela e-IT ngengqiqo yexesha langempela kunye nohlalutyo oluqikelelweyo, oluvumela ukwenza izigqibo eziqhutywa yidatha.

IziCwangciso zokuPhumeza

Ukuphumeza ii-AIOps kufuna indlela yobuchule:

  1. Ukuqokelelwa kwedatha: Qokelela idatha ye-IT kwimithombo eyahlukeneyo, kubandakanywa iilog, iimetrics, kunye neefayile zoqwalaselo.
  2. Udityaniso lweDatha: Dibanisa idatha esuka kwimithombo eyahlukeneyo kwiqonga elinye.
  3. **Ukufunda ngomatshini **: Faka i-algorithms yokufunda ngomatshini ukuhlalutya idatha kunye nokuqikelela imiba enokubakho ye-IT.
  4. **I-Automation **: Yenza imisebenzi yesiqhelo kunye nokuhamba komsebenzi.
  5. **Ukubeka iliso kunye neNgxelo **: Ukubeka iliso ngokuqhubekayo ukusebenza kwe-AIOps kunye nokubonelela ngengxelo ukuphucula ukuchaneka nokusebenza.

Izifundo zeMyelo zeHlabathi yokwenyani

Imibutho eliqela iphumeze ngempumelelo ii-AIOps ukuguqula ukuhamba komsebenzi wolawulo lwe-IT:

  1. I-Verizon: I-Verizon iphunyezwe i-AIOps ukuphucula ukuthembeka kwenethiwekhi kunye nokunciphisa ixesha lokuphumla. Iqonga lisebenzise ukufundwa komatshini ukuqikelela nokuthintela imiba yenethiwekhi, okukhokelela ekunciphiseni kwe-30% ngexesha lokuphumla.
  2. ** I-AT & T **: I-AT & T iphunyezwe i-AIOps ukuphucula ulawulo lweziganeko kunye nokunciphisa ixesha eliqhelekileyo lokubona (MTTD) kunye nexesha elichanekileyo lokusombulula (MTTR). Iqonga lisebenzise i-automation ukufumanisa nokusombulula iziganeko ze-IT, okubangelwa ukunciphisa i-40% kwi-MTTD kunye nokunciphisa i-50% kwi-MTTR.
  3. I-Microsoft: I-Microsoft iphumeze i-AIOps ukuphucula ucazululo lokuxela kwangaphambili kunye ne-automation. Iqonga lisebenzise ukufundwa komatshini ukuqikelela imiba enokubakho ye-IT, ivumela ukugcinwa okusebenzayo kunye nokunciphisa ixesha lokuphumla.

Code Block: AIOps Umzekelo kwiPython

Nantsi ibhloko yekhowudi entsonkothileyo kwiPython ebonisa ukusetyenziswa kwe-AIOps kulawulo lwe-IT:

ngenisa iipandas njenge pd
ukusuka sklearn.ensemble yokungenisa RandomForestClassifier
ukusuka sklearn.model_selection yokungenisa train_test_split
ukusuka sklearn.metrics import accuracy_score

# Layisha idatha
idatha = pd.read_csv('it_data.csv')

# Lungiselela idatha
X = data.drop(['isiganeko'], i-axis=1)
y = idatha['isiganeko']

# Yahlula idatha kwiiseti zoqeqesho kunye novavanyo
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Qeqesha udidi lwehlathi olungakhethiyo
rfc = RandomForestClassifier(n_estimators=100, random_state=42)
rfc.fit(X_train, y_train)

# Yenza uqikelelo
y_pred = rfc.predict(X_test)

# Vavanya imodeli
ukuchaneka = accuracy_score(y_test, y_pred)
print(f'Ukuchaneka: {chaneka:.2f}')

# Sebenzisa imodeli ukuqikelela imiba ye-IT enokubakho
data_entsha = pd.DataFrame({'cpu_usage': [80], 'memory_usage': [90]})
uqikelelo = rfc.predict(idatha_entsha)
ukuba uqikelelo[0] == 1:
    shicilela('Ingxaki ye-IT enokubakho ichongiwe!')

Le khowudi yekhowudi isebenzisa i-classifier yehlathi engahleliweyo ukuqikelela imiba enokubakho ye-IT esekelwe kwi-CPU kunye nedatha yokusetyenziswa kwememori. Imodeli iqeqeshwe kwiinkcukacha zembali kwaye isetyenziselwa ukwenza uqikelelo kwidatha entsha.

Umboniso: i-AIOps Platform Architecture

Nantsi imbonakalo yeqonga loyilo lwe-AIOps usebenzisa iChart.js:

{
  "type": "pieChart",
  "idatha": {
    "labels": ["Ukuqokelelwa kweDatha", "uNxibelelwano lweDatha", "Ukufunda ngoomatshini", "Ukuzenzekela", "UkuJonga kunye neNgxelo"],
    "iiseti zedatha": [
      {
        "ileyibhile": "I-AIOps Platform Architecture",
        "idatha": [30, 20, 20, 15, 15],
        "backgroundColor": ["#FF6384", "#36A2EB", "#FFCE56", "#4BC0C0", "#FF99CC"],
        "hoverBackgroundColor": ["#FF99CC", "#36A2EB", "#FFCE56", "#4BC0C0", "#FF99CC"]
      }
    ]
  },
  "iinketho": {
    "isihloko": {
      "bonisa": yinyani,
      "umbhalo": "AIOps Platform Architecture"
    },
    "legend": {
      "bonisa": yinyani
    }
  }
}

Lo mzobo ubonisa amacandelo ahlukeneyo oyilo lweqonga le-AIOps, kubandakanywa ukuqokelela idatha, ukuhlanganiswa kwedatha, ukufundwa komatshini, ukuzenzekelayo, kunye nokubeka iliso kunye nempendulo.

Isiphelo

I-AIOps iteknoloji yokutshintsha umdlalo oguqula indlela amaqela e-IT alawula ngayo kwaye agcine iinkqubo ze-IT ezinzima. Ngokusebenzisa ukufundwa koomatshini kunye ne-AI, i-AIOps inika amandla ulawulo lwe-IT olusebenzayo, uhlalutyo lokuxela kwangaphambili, kunye ne-automation, ekhokelela ekuphuculeni ukusebenza kakuhle, ukuthembeka okuphuculweyo, kunye nokwanda kwemveliso. Ngokuphumeza ii-AIOps, imibutho inokunciphisa ixesha lokuphumla, iphucule ukwenziwa kwezigqibo, kwaye izuze umda wokukhuphisana kwimarike.

Njengoko itekhnoloji iqhubeka nokuvela, sinokulindela ukubona ngakumbi usetyenziso lwe-AIOps kulawulo lwe-IT. Ngamandla ayo okuguqula indlela esilawula ngayo iinkqubo ze-IT, i-AIOps yintsimi enika umdla nekhula ngokukhawuleza ekufanele ukuba ibekwe iliso.

Have Questions?

Let's Talk

Contact me on LinkedIn